IJACSA Volume 13 Issue 8 - thesai.org

IJACSA Volume 13 Issue 8

Copyright Statement: This is an open access publication licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, even commercially as long as the original work is properly cited.

View Full Issue

Paper 1: Real-Time Wildfire Detection and Alerting with a Novel Machine Learning Approach

Abstract: Up until the end of July 2022, there have been over 38k wildfires in the US alone, decimating over 5.6 million acres. Wildfires significantly contribute to carbon emission, which is the root cause of global warming. Research has shown that artificial intelligence already plays a very important role in wildfire management, from detection to remediation. In this investigation a novel machine learning approach has been defined for spot wildfire detection in real time with high accuracy. The research compared and examined two different Convolutional Neural Network (CNN) approaches. The first approach; a novel machine learning method, a model server framework is used to serve convolutional neural network models trained for daytime and nighttime to validate and feed wildfire images sorted by different times of day. In the second approach that has been covered by existing research, one big CNN model is described as training all wildfire images regardless of daytime or nighttime. With the first approach, a higher detection precision of 98% has been achieved, which is almost 8% higher than the result from the second approach. The novel machine learning approach can be integrated with social media channels and available forest response systems via API’s for alerting to create an automated wildfire detection system in real time. This research result can be extended by fine tuning the CNN network model to build wildfire detection systems for different regions and locations. With the rapid development of network coverage such as Starlink and drone surveillance, real time image capturing can be combined with this research to fight the increasing risk of wildfires with real time wildfires detection and alerting in automation.

Author 1: Audrey Zhang

Author 2: Albert S. Zhang

Keywords: Wildfire detection; CNN (convolutional neural network); machine learning; image processing; model server framework

Paper 2: Severely Degraded Underwater Image Enhancement with a Wavelet-based Network

Abstract: Underwater images are important in marine science and ocean engineering fields owing to giving color information, low cost, and compact. Yet obtained underwater images are often degraded and restoring and enhancing wavelength selective signal attenuation of underwater images depending on complex underwater physical process is essential in practical application. While recently developed deep learning is a promising choice, constructing sufficiently large dataset covering whole real images is challenging, peculiar to underwater image processing. In order to supplement relatively small dataset, previous studies alternatively construct an artificial underwater image dataset based on a physical model or Generative Adversarial Network. Also, incorporating traditional signal processing methods into the network architecture has shown promising success, though enhancement of severely degraded underwater images remains to be a big issue. In this paper, we tackle underwater image enhancement based on an encoder-decoder based deep learning model incorporating discrete wavelet transform and whitening and coloring transform. We also construct a severely degraded real underwater image dataset. The presented model shows excellent results both qualitatively and quantitatively in the artificial and real image dataset. Constructed dataset is available at https://github.com/tkswalk/2022-IJACSA.

Author 1: Shunsuke Takao

Author 2: Tsukasa Kita

Author 3: Taketsugu Hirabayashi

Keywords: Underwater image enhancement; deep learning; discrete wavelet transform; whitening and coloring transform

Paper 3: Towards Personalized Adaptive Learning in e-Learning Recommender Systems

Abstract: An adaptive e-learning scenario not only allows people to remain motivated and engaged in the learning process, but it also helps them expand their awareness of the courses they are interested in. e-Learning systems in recent years had to adjust with the advancement of the educational situation. Therefore many recommender systems have been presented to design and provide educational resources. However, some of the major aspects of the learning process have not been explored quite enough; for example, the adaptation to each learner. In learning, and in a precise way in the context of the lifelong learning process, adaptability is necessary to provide adequate learning resources and learning paths that suit the learners’ characteristics, skills, etc. e-Learning systems should allow the learner to benefit the most from the presented learning resources content taking into account her/his learning experience. The most relevant resources should be recommended matching her/his profile and knowledge background not forgetting the learning goals she/he would like to achieve and the spare time she/he has in order to adjust the learning session with her/his goals whether it is to acquire or reinforce a certain skill. This paper proposes a personalized e-learning system that recommends learning paths adapted to the users profile.

Author 1: Massra Sabeima

Author 2: Myriam Lamolle

Author 3: Mohamedade Farouk Nanne

Keywords: e-Learning; adaptive learning; recommendation system; ontology

Paper 4: A Comparative Research on Usability and User Experience of User Interface Design Software

Abstract: With the development of science and technology, people increasingly rely on intelligent interactive products, thus promoting the vigorous development of the user interface industry. Software with high usability and user experience can improve users’ effectiveness and satisfaction, as well as the user viscosity. Taking three design software: Sketch, Adobe XD, and Figma, which is most frequently used by design industry practitioners and students, as research cases, this study compared and discussed the impact of interaction design and interface layout on the usability and user experience combining with subjective experiment methods, scale scoring, user testing and retrospective think-aloud interview, as well as objective experiment method, eye tracking. It is found that the overall usability and user experience of Figma is the best, Adobe XD is the second, and Sketch is the worst. The main reason for this result is that the three software have different degrees of issues in interface layout, information quality, and interaction logic. Based on the results, the optimization suggestions for the usability and user experience of user interface design software are proposed from three perspectives: interface design, information quality and interaction design.

Author 1: Junfeng Wang

Author 2: Zhiyu Xu

Author 3: Xi Wang

Author 4: Jingjing Lu

Keywords: Usability; user experience; interaction design; UI design software; eye tracking

Paper 5: Experimental Evaluation of Basic Similarity Measures and their Application in Visual Information Retrieval

Abstract: Searching for similar images is an important feature for image databases and decision support systems in various subject domains. However, it is essential that search results are sorted by degree of similarity in reverse order. This paper presents a comparative analysis of four existing similarity measures and experimentally tests whether they could be used to calculate similarity between images. Metrics could be evaluated by comparing their results to the cumulative human perception of similarity between the same images, obtained by real people. However, this introduces a lot of subjectivism due to non-uniform judgement and evaluation scales. The paper presents a more objective approach - checks which measure performs best in retrieving more images, containing objects of the same type. Results show all four measures could be used to calculate similarity between images, but Jaccard’s index performs best in most cases, because it compares features vectors positionally and thus indirectly consider shape, position, orientation and other features.

Author 1: Miroslav Marinov

Author 2: Yordan Kalmukov

Author 3: Irena Valova

Keywords: Content Based Image Retrieval (CBIR); image search and ranking; similarity measures; image databases

Paper 6: Automatic Detection of Roads using Aerial Photographs and Calculation of the Optimal Overflight Route for a Fixed-wing Drone

Abstract: Currently, fixed-wing drones have become indispensable tools for the surveillance of large areas of land, justified by their better cost/benefit ratio, great flight autonomy, and payload capacity. In particular, the identification of roads, traffic control, monitoring of wear on asphalt layers, risk identification, and safety improvement are applications that are being implemented in these unmanned aerial vehicles. Tracking a road requires systems capable of detecting artificial marks through images employing aerial photographs that allow the implementation of optimal overflight routes. This research work presents a solution to the problem of road tracking from aerial photographs and implements an image processing algorithm and morphological techniques that calculate and traces the ideal route for the drone to track automatically, regardless of its orientation and the type of road.

Author 1: Miguel Pérez P

Author 2: Holman Montiel A

Author 3: Fredy Martínez S

Keywords: Automatic road tracking; decorrelation stretching; aerial imagery; optimal overflight route calculation; UAV

Paper 7: Research on Regional Differentiation Allocation Mode of Energy Finance based on Attention Mechanism and Support Vector Machine

Abstract: This paper studies the prediction method of regional differentiated allocation mode of energy finance based on attention mechanism and support vector machine to provide scientific guidance for the future development direction of energy finance in each region. Analysis of the key factors influencing the energy consumption, through the attention mechanism to extract the regional factors such features constitute the details of the sample set, the characteristics of the sample set after implementation of fusion and normalized processing, gain new characteristics of sample set as input to construct support vector machine forecasting model, prediction of energy consumption in each region of the output. According to the results, the differentiated allocation patterns of energy finance in each region are predicted. The results show that the prediction model of this method has high training and test prediction accuracy, and the prediction results are consistent with the actual data in historical statistics. Compared with the existing methods, the method of this study can more scientifically and effectively predict the sustainable and stable development of energy finance in various regions of the city in the future. The energy consumption of the experimental city predicted in this study in the next nine years is from high to low in the order of region C, region a and region B. from this, it is predicted that the regions A, B and C of this city in the future will be applicable to the government market dual oriented Government oriented and market-oriented energy finance allocation models. The prediction results can provide scientific guidance for the sustainable and stable development of energy finance in various regions of the city in the future.

Author 1: Ling Sun

Author 2: Hao Wu

Keywords: Attention mechanism; support vector machine; energy finance; differentiation configuration mode; energy consumption

Paper 8: Bilingual AI-Driven Chatbot for Academic Advising

Abstract: Conversational technologies are revolutionizing how organizations communicate with people, thereby raising quick responses and constant availability expectations. Students often have queries about the institutional and academic policies and procedures, academic progression, activities, and more in an academic environment. In reality, the student services team and the academic advisors are overwhelmed with several queries that they cannot provide instant responses to, resulting in dissatisfaction with services. Our study leverages Artificial Intelligence and Natural Language processing technologies to build a bilingual chatbot that interacts with students in the English and Arabic languages. The conversational agent is built in Python and designed for students to support advising-related queries. We use a purpose-built domain-specific corpus consisting of the common questions advisors receive from students and their responses as the chatbots knowledge base. The chatbot engine determines the user intent by processing the input and retrieves the most appropriate response that matches the intent with an accuracy of 80% in English and 75% in Arabic. We also evaluated the chatbot interface by conducting field experiments with students to test the accuracy of the chatbot with real-time input and test the application interface.

Author 1: Ghazala Bilquise

Author 2: Samar Ibrahim

Author 3: Khaled Shaalan

Keywords: Chatbot; conversational agent; academic advising; natural language processing; deep learning; bilingual English Arabic

Paper 9: Modified Prophet+Optuna Prediction Method for Sales Estimations

Abstract: A prediction method for estimation of sales based on Prophet with a consideration of nonlinear events and conditions by a modified Optuna is proposed. Linear prediction does not work for a long-term sales prediction because purchasing actions are based on essentially nonlinear customers’ behavior. One of nonlinear prediction methods is the well-known Prophet. It, however, is still difficult to adjust the nonlinear parameters in the Prophet. To adjust the parameters, the Optuna is widely used. It, however, is not good enough for parameter tuning by the Optuna. Therefore, the Optuna is modified with a short-term moving mean and standard deviation of the sales for final prediction. More than that, specific event such as typhoon event is to be considered in the sales prediction. Through experiments with a real sales data, it is found the sensitivity of the parameters the upper window, lower window, event dates, etc. for the final sales and the effect of the Optuna is 11.73%. Also, it is found that the effect of the consideration of Covid-19 is about 2.4% meanwhile the effect of the proposed modified Optuna is around 3 % improvement of the prediction accuracy (from 80 % to 83 %).

Author 1: Kohei Arai

Author 2: Ikuya Fujikawa

Author 3: Yusuke Nakagawa

Author 4: Tatsuya Momozaki

Author 5: Sayuri Ogawa

Keywords: Prediction method; nonlinearity; prophet; optuna; typhoon event; modified optuna; mean and standard deviation adjustment

Paper 10: Encryption Algorithms Modeling in Detecting Man in the Middle Attack in Medical Organizations

Abstract: A cyberattack is a serious crime that could affect medical organizations. These attacks could affect medical organization sensitive data disclosure, loss of organization data, or the business's continuity. The Man-in-The-Middle (MITM) attack is one of the threats that could impact medical organizations. It happens when unapproved outsiders break into the traffic between two parties that think they are conversing directly. At the same time, the adversary can access, read, and change secret information. Because of that, medical organizations lose confidentiality, integrity, and availability. Data encryption is a solution that changes vital information to unreadable by unauthorized and unintended parties. It could involve protecting data with cryptography, usually by leveraging a scrambled code. Only the individuals with the decoding key can read the information. There is no full protection due to the variety of MITM attacks. Each encryption algorithm has its advantages and disadvantages, like the speed of encryption and decryption, strength of the algorithm, and the cipher type. This research investigates the MITM attacks and comprehensively compares the Rivest Shamir Adleman algorithm and the Elliptic Curve Cryptography algorithm.

Author 1: Sulaiman Alnasser

Author 2: Raed Alsaqour

Keywords: Cyberattack; medical organization; man in the middle attack; encryption algorithm; rivest shamir adleman algorithm; elliptic curve cryptography algorithm

Paper 11: A Smart Decision Making System for the Optimization of Manufacturing Systems Maintenance using Digital Twins and Ontologies

Abstract: Now-a-days manufacturing processes are becoming more and more complex which constantly complicate the management of their life cycle. Although, in order to survive and maintain a good position in the competitive industrial context, industrials have understood that they must optimize the whole life cycle of their manufacturing processes. The maintenance constitutes one of the key processes indispensable to ensure the proper functioning and to optimize the lifetime of machines and production lines, and thus to optimize quality and production costs. Therefore, its automation and optimization represent until now a center of interest for researches and manufacturers, especially those related to the integration of artificial intelligence tools in the industry. In this context, several new concepts and technologies have emerged, particularly in the context of industry 4.0. One of these new concepts is digital twins, which has become a promising direction to optimize manufacturing processes lifecycle. However, the implementation of this technology faces several complex problems related to the interoperability between physical entities and their virtual counterparts, as well as to the logical reasoning between the different elements constituting the digital twin. It is in this context that an approach based on digital twins and ontologies is proposed. The originality of this paper lies in two important points: the first is the exploitation of the expressiveness and reasoning capabilities of ontologies to solve cyber-physical interoperability problems at the digital twin level, while the second is the automation of the whole maintenance process and its decision making key points using the inference potentialities of ontologies. The applicability and effectiveness of the proposed approach is validated through an industrial case of study.

Author 1: ABADI Mohammed

Author 2: ABADI Chaimae

Author 3: ABADI Asmae

Author 4: BEN-AZZA Hussain

Keywords: Maintenance systems; maintenance policy; digital twin; reasoning; ontologies; automation; cyber-physical interoperability; decision making; artificial intelligence

Paper 12: An Improved Arabic Sentiment Analysis Approach using Optimized Multinomial Naïve Bayes Classifier

Abstract: Arabic sentiment analysis has emerged during the last decade as a computational process on Arabic texts for extracting people's attitudes toward targeted objects or their feelings and emotions regarding targeted events. Sentiment analysis (SA) using machine learning (ML) methods has become an important research task for developing various text-based applications. Among different ML classifiers, multinomial Naïve Bayes (MNNB) classifier is widely used for documents classification due to its ability for performing statistical analysis of text contents. It significantly simplifies textual-data classification and offers an alternative to heavy ML-based semantic analysis methods. However, the MNNB classifier has a number of hyper-parameters affects the classification task of texts and controls the decision boundary of the model itself. In this paper, an optimized MNNB classifier-based approach is proposed for improving Arabic sentiment analysis. A number of experiments are conducted on large sets of Arabic tweets to evaluate the proposed approach. The optimized MNNB classifier is trained on three datasets and tested on a different separated test set to show the performance of developed approach. The experimental results on the test set revealed that the optimized MNNB classifier of proposed approach outperforms the traditional MNNB classifier and other baseline classifiers. The accuracy rate of the optimization approach is increased by 1.6% compared with using the default values of the classifier’s hyper-parameters.

Author 1: Ahmed Alsanad

Keywords: Machine learning; Arabic sentiment analysis; optimized multinomial Naïve Bayes (MNNB) classifier; hyper-parameters optimization

Paper 13: Erratic Navigation in Lecture Videos using Hybrid Text based Index Point Generation

Abstract: The difficulty in lecture videos is an erratic navigation in lecture video for watching only the needed portion of video content. Machine learning technologies like Optical Character Recognition and Automatic Speech Recognition allows to easily fetch the information that is hybrid text from lecture slides and audio respectively. This paper presents three main analysis for hybrid text retrieval, which is further useful for indexing the video. The experimental results indicate that the key frame extraction accuracy is 94 percent. The accuracy of the Slide-To-Text conversion achieved by this study's evaluation of the text extraction capability of Tesseract, Abbyy Finereader, Transym, and the Google Cloud Vision Optical Character Recognition is 92.0%, 90.5%, 80.8%, and 96.7% respectively. Similarly the result of title identification is about 96 percent. To extract the speech text three different APIs are used namely, Microsoft, IBM, and Google Speech-to-Text API. The performance of the transcript generator is measured using Word Error Rate, Word Recognition Rate, and Sentence Error Rate metrics. This paper found that Google Cloud Vision Optical Character Recognition and Google Speech-to-Text API have achieved best results compared to other methods. The results obtained are very good and agreeable, therefore the proposed methods can be used for automating the lecture video indexing.

Author 1: Geeta S Hukkeri

Author 2: R. H. Goudar

Keywords: Automatic speech recognition; indexing; key-frames; lecture video; optical character recognition; title identification; text extraction

Paper 14: High Capacity Image Steganography System based on Multi-layer Security and LSB Exchanging Method

Abstract: Data security is becoming an important issue because of the vast use of the Internet and data transfer from one place to another. Security of these data is essential, especially when these data represent critical information. There are several techniques used to hide these data, such as encryption. Steganography can be utilised as an alternative to encryption because encryption is susceptible to data modification during transmission. Steganography is the hiding data on a cover multimedia such as images, audio, and video. The technique allows security for data transmission so unwanted third parties cannot notice the hidden data. The challenge of steganography is the trade-off between the hidden data's payload capacity and the system's imperceptibility and robustness. If the hidden data increases, the imperceptibility and the robustness will be decreased. This case is a big challenge in this digital world where social media, Internet, and data transfer are used hugely. Because of this, this paper proposes using a modified Least Significant Bit (LSB) method for the embedding process called Multi-Layer Least Significant Bit Exchange Method (MLLSBEM). This proposed algorithm uses the AES encryption method to encrypt the secret text and then uses Huffman coding to compress the encrypted message as pre-processing data. The proposed study seeks to strike a compromise between important issues, provide maximum payload capacity, and retain high security, imperceptibility, and reliability for secret communication Using image processing and steganography techniques. Simulation findings demonstrate that the suggested method is superior for existing PSNR, SSIM, NCC, and payload capacity investigations. The proposed method is immune to the histogram, chi-square, and HVS attacks.

Author 1: Rana Sami Hameed

Author 2: Siti Salasiah Mokri

Author 3: Mustafa Sabah Taha

Author 4: Mustafa Muneeb Taher

Keywords: Information hiding; steganography; cryptography; multi-layer security; high capacity component

Paper 15: Recognition of Odia Character in an Image by Dividing the Image into Four Quadrants

Abstract: This paper deals with optical character recognition of Odia characters written in a particular font family ‘AkrutiOriAshok-99’ with different font sizes 18, 20, 22, 24, 26, 28, 36, 48 and 72 in Bold style. The font ‘AkrutiOriAshok-99’ is a font from the typing software ‘Akruti’. The basic idea behind the approach followed in this paper is the character decomposition into four quadrants and then extracting features from each quadrant. The image processing techniques like converting the image to gray, resizing of image and converting gray image to binary are used in this approach. The system explained in this paper has two major parts: DictionaryBuilding and FindingMatch. For DictionaryBuilding, dictionary of images which are created either by scanning a document or a document converted to image, both written in same font family in different sizes. The features are extracted from each image in any font size in the ‘Dictionary’ using Preprocessing, FindPath, GettingFeaturesLeft or GettingFeaturesRight, VisitSubQuad, RemainingSubQuad, WriteToExcel and CommonFeature modules. The part FindingMatch is responsible for finding a correct match in the dictionary for the input image. For this, FeatureExtraction and Recognition modules have been used. Longest Common Subsequence (LCS) has been used for finding the common feature in DictionaryBuilding as well as finding the correct match. A total of 1800 characters, 200 characters of each font size have been tested and 98.1% of correctness has been achieved.

Author 1: Aradhana Kar

Author 2: Sateesh Kumar Pradhan

Keywords: Odia characters; image processing; character decomposition; machine learning; optical character recognition

Paper 16: Enhancement of Design Level Class Decomposition using Evaluation Process

Abstract: Refactoring on the design level artifact such as the class diagram was already done using the threshold-based agglomerative hierarchical clustering method, specifically class decomposition. The approach produced a better cluster based on the label name similarity of attribute and method. But, some problems emerge from the experiment result. The negative Silhouettes element still exist in the cluster. And, there is an unusable cluster that only consists of one attribute element. This paper has proposed the evaluation process to optimize the result of clustering. This evaluation process is an additional process that aims to move the negative Silhouettes element to the other cluster. The movement is also to get the better value of element Silhouettes value. The evaluation process can produce a better result for clusters. The clusters produced from the evaluation process have higher Silhouettes values. The average Silhouettes value is increased by about 40%. Ultimately, the result shows no unusable cluster as mentioned in the previous research.

Author 1: Bayu Priyambadha

Author 2: Tetsuro Katayama

Keywords: Refactoring; design level refactoring; software refactoring; class decomposition; software quality

Paper 17: A Multi-Objective Optimization for Supply Chain Management using Artificial Intelligence (AI)

Abstract: Supply chain management seeks to solve the complex problems of transporting goods from the suppliers to the end customers. Improving the differentiation between different paths to reduce costs and time may require smart systems. This paper proposes two new algorithms for determining, with Multi-Objective Optimization, the least cost and the most appropriate path between two nodes. First: Ant colony optimization (ACO) algorithm, working alongside with Multi Objective Optimization (MOO), is adopted to determine the shortest path and time between two nodes to reach with the least cost. Multi-Objective intelligent Ant Colony (MOIAC) algorithm improves supply chain management to achieve the optimal and the most appropriate solutions. Second: Particle Swarm Optimization (PSO) algorithm, also working alongside MOO, is adopted to determine the least cost, time, and shortest path. Multi Optimization Intelligent Particle Swarm (MOIPS) algorithm improves supply chain management by determining the shortest path with the least cost. These two proposed algorithms seek the optimal solution by MOO using a JAVA Program. The experimental results show the excellence of the first algorithm in determining the optimal and the most appropriate path while getting throw risks inherent in transporting goods. It also demonstrates excellence in transporting goods in the shortest possible time and with the least cost. The second algorithm also shows excellence in transporting goods with the least possible cost via the shortest path and in the shortest time.

Author 1: Mohamed Hassouna

Author 2: Ibrahim El-henawy

Author 3: Riham Haggag

Keywords: Supply chain management; artificial intelligence; particle swarm optimization; ant colony optimization and multi-objective optimization

Paper 18: Cylinder Liner Defect Detection and Classification based on Deep Learning

Abstract: The machine vision-based defect detection for cylinder liner is a challenging task due to irregular shape, various and small defects on the cylinder liner surface. To improve the accuracy of defect detection by machine vision a deep learning-based defect detection method for cylinder liner was explored in this paper. First, a machine vision system was designed based on the analysis of the causes and types of defects to obtain the field images for establishing an original dataset. Then the dataset was augmented by a modified augmentation method which combines the region of interest automatic extraction method with the traditional augmentation methods. Except for introduction of the anchor configuration optimization method, an XML file-based method of highlighting defect area was proposed to address the problem of tiny defect detection. The optimal model was experimentally determined by considering the network model, the training strategy and the sample size. Finally, the detection system was developed and the network model was deployed. Experiments are carried out and the results of the proposed method compared with those of the traditional methods. The results show that the detection accuracies of sand, scratch and wear defects are 77.5%, 70% and 66.3% which are improved by at least 26.3% compared with the traditional methods. The proposal can be used for field defect detection of cylinder liner.

Author 1: Chengchong Gao

Author 2: Fei Hao

Author 3: Jiatong Song

Author 4: Ruwen Chen

Author 5: Fan Wang

Author 6: Benxue Liu

Keywords: Cylinder liner; defect detection; deep learning; machine vision

Paper 19: Mobile Payment Transaction Model with Robust Security in the NFC-HCE Ecosystem with Secure Elements on Smartphones

Abstract: The Near Field Communication embedded (NFC-embedded) smartphone consists of two ecosystems, namely Near Field Communication Subscriber Identity Module Secure Element (NFC-SIM-SE) and Near Field Communication Host Card Emulation (NFC-HCE). NFC-SIM-SE places secure elements in smartphones, while NFC-HCE places secure elements in the cloud. In terms of security, the location of secure elements in the cloud is one of the weaknesses of NFC-HCE. The APL-SE transaction model is developed as a solution to improve transaction security with NFC-enabled mobile. This model moves the secure elements of the NFC-HCE ecosystem from the cloud to the smartphone so that when the transaction is made, the smartphone does not communicate with the outside network to access the secure element. The APL-SE transaction model is tested using dummy data to calculate the processing time measurements for each step. The model is also tested for the encryption process. The encrypted data is compared with the original data, then the randomness is calculated. This transaction model is also tested by looking at the data randomness, which shows that the encrypted data is declared random. Random data increases data security. The transaction model test shows that the transaction runs well because the encrypted data is proven random, and the execution time is 1,074 ms. The time of 1,074 ms is far below an attacker's time to decipher the encrypted data. Random and fast encryption results indicate that transactions are secure. This achievement makes the opportunity for attackers to manipulate data small, so security is increased.

Author 1: Lucia Nugraheni Harnaningrum

Author 2: Ahmad Ashari

Author 3: Agfianto Eko Putra

Keywords: Transaction; near field communication; mobile; secure element; encryption

Paper 20: Observation of Imbalance Tracer Study Data for Graduates Employability Prediction in Indonesia

Abstract: Tracer Study is a mandatory aspect of accreditation assessment in Indonesia. The Indonesian Ministry of Education requires all Indonesia Universities to anually report graduate tracer study reports to the government. Tracer study is also needed by the University in evaluating the success of learning that has been applied to the curriculum. One of the things that need to be evaluated is the level of absorption of graduates into the working industry, so a machine learning model is needed to assist the University Officials in evaluating and understanding the character of its graduates, so that it can help determine curriculum policies. In this research, the researcher focuses on making a reliable machine learning model with a tracer study dataset format that has been determined by the Government of Indonesia. The dataset was obtained from the tracer study of Amikom University. In this study, SVM will be tested with several variants of the algorithm to handle imbalanced data. The study compared SMOTE, SMOTE-ENN, and SMOTE-Tomek combined with SVM to detect the employability of graduates. The test was carried out with K-Fold Cross Validation, with the highest accuracy and precision results produced by SMOTE-ENN SVM model by value of 0.96 and 0.89.

Author 1: Ferian Fauzi Abdulloh

Author 2: Majid Rahardi

Author 3: Afrig Aminuddin

Author 4: Sharazita Dyah Anggita

Author 5: Arfan Yoga Aji Nugraha

Keywords: Tracer study; support vector machine; synthetic minority oversampling technique; SMOTE; employability

Paper 21: Determining the Best Email and Human Behavior Features on Phishing Email Classification

Abstract: There are many email filters that have been developed for classifying spam and phishing email. However, there is still a lack of phishing email filters developed because of the complexity of feature extraction and selection of the data. There are several categories of features for classifying phishing emails, either on the email part or on the human part. The absence of which features are best for helping to classify phishing emails is one of the challenges; in the previous experiment, there was no benchmark for the features to be used for phishing email classification. This research will provide new insight into the feature selection process in the phishing email classification area. Therefore, this work extracts the features based on the category and determines which features have the most impact on classifying email as phishing or not phishing using a machine learning approach. Feature selection is one of the essential parts of getting a good classification result. Therefore, obtaining the best features from email and human behavior will significantly impact phishing classification. This research collects the public phishing email dataset, extracts the features based on category using Python, and determines the feature importance using machine learning approaches with the PyCaret library. The dataset experimented on three different experiments in which each feature category was separated, and one experiment was the combined feature selection. Binary classification is also done with the extracted features. The experiment verified that the proposed method gave a good result in feature importance and the binary classification using selected features in terms of accuracy compared to previous research. The highest result obtained is the classification with combined features with 98% accuracy. The results obtained are better compared to previous studies. Hence, this research proves that the selected features will increase the performance of the classification.

Author 1: Ahmad Fadhil Naswir

Author 2: Lailatul Qadri Zakaria

Author 3: Saidah Saad

Keywords: Phishing; phishing email classification; features selection; binary classification; email features; human features

Paper 22: A Scalable Machine Learning-based Ensemble Approach to Enhance the Prediction Accuracy for Identifying Students at-Risk

Abstract: Among the educational data mining problems, the early prediction of the students' academic performance is the most important task, so that timely and requisite support may be provided to the needy students. Machine learning techniques may be used as an important tool for predicting low-performers in educational institutions. In the present paper, five single-supervised machine learning techniques have been used, including Decision Tree, Naïve Bayes, k-Nearest-Neighbor, Support Vector Machine, and Logistic Regression. To analyze the effect of an imbalanced dataset, the performance of these algorithms has been checked with and without various resampling methods such as Synthetic Minority Oversampling Technique (SMOTE), Borderline SMOTE, SVM-SMOTE, and Adaptive Synthetic (ADASYN). The Random hold-out method and GridSearchCV were used as model validation techniques and hyper-parameter tuning respectively. The results of the present study indicated that Logistic Regression is the best performing classifier with every balanced dataset generated using all of the four resampling techniques and also achieved the highest accuracy of 94.54% with SMOTE. Furthermore, to improve the prediction results and to make the model scalable, the most suitable classifier was integrated with the help of bagging, and a well-accepted accuracy of 95.45% was achieved.

Author 1: Swati Verma

Author 2: Rakesh Kumar Yadav

Author 3: Kuldeep Kholiya

Keywords: Educational data mining; resampling methods; feature selection technique; machine learning; imbalanced data

Paper 23: AI-based Academic Advising Framework: A Knowledge Management Perspective

Abstract: Academic advising has become a critical factor of students’ success as universities offer a variety of programs and courses in their curriculum. It is a student-centered initiative that fosters a student’s involvement with the institution by supporting students in their academic progression and career goals. Managing the knowledge involved in the advising process is crucial to ensure that the knowledge is available to those who need it and that it is used effectively to make good advising decisions that impact student persistence and success. The use of AI-based tools strengthens the advising process by reducing the workload of advisors and providing better decision support tools to improve the advising practice. This study explores the challenges associated with the current advising system from a knowledge management perspective and proposes an integrated AI-based framework to tackle the main advising tasks.

Author 1: Ghazala Bilquise

Author 2: Khaled Shaalan

Keywords: Knowledge management; artificial intelligence; academic advising; rule-based expert system; machine learning; chatbot; conversational agent

Paper 24: An Adaptation Layer for Hardware Restrictions of Quadruple-Level Cell Flash Memories

Abstract: In recent years, major flash memory vendors have produced SSDs and fusion memories as substitution for hard disks. However, there has been a lack of studies on access restriction of QLC flash memory, since most researches have targeted small capacity flash memory. As a solution, we propose to implement an adaptation layer between the file system and FTL (Flash Translation Layer). Instead of immediately writing data given from file system to flash memory, the adaptation layer gathers and adjusts data in the unit of a page, and separates random data from sequential data. By implementing the adaptation layer, previous FTL algorithms can be fully applied on the QLC flash memory. According to our experiment, the adaptation layer forms smaller number of pages than the current data gathering algorithm.

Author 1: Se Jin Kwon

Keywords: Cache storage; flash memory; SSD; nonvolatile memory

Paper 25: Improving Internet of Things Platform with Anomaly Detection for Environmental Sensor Data

Abstract: Internet of things has an essential role in various application domains. The number of Internet of Things applications makes researchers try to formulate how to design the architecture of the Internet of Things platform so that it can be used generically in various domains. Commonly used architectural designs consist of data collecting, data preprocessing, data analysis, and data visualization. However, sensor data that enters the platform often experiences anomalies such as constant values or being stuck-at zero, which are processed manually at the data preprocessing stage. In this research, we try to design an anomaly detection system on the Internet of Things platform that can automatically improve the platform's performance in detecting anomalies. In this study, we compared the False Positive Rate of several anomaly detection algorithms tested to real datasets in the environmental sensor data application domain. The results showed that the anomaly detector system on the Internet of Things platform had an optimal False Positive Rate of 0.9%.

Author 1: Okyza Maherdy Prabowo

Author 2: Suhono Harso Supangkat

Author 3: Eueung Mulyana

Author 4: I Gusti Bagus Baskara Nugraha

Keywords: Anomaly detection; sensor data; multivariate; Internet of Things; smart system

Paper 26: Math Balance Aids based on Internet of Things for Arithmetic Operational Learning

Abstract: Industry 4.0 has changed various aspects of human life towards an era heavily influenced by information technology. The impact of industry 4.0 on the education sector has led to the emergence of the term education 4.0. The Internet of Things (IoT) is one of the pillars of industry 4.0. With its capabilities, IoT can provide opportunities to develop innovations in the field of education. Several studies show that teaching aids can improve the quality of learning and learning outcomes. In Indonesia, mathematics is a compulsory subject taught from elementary school to higher education. Previous studies that used mathematical (math) balance aids to help students learn mathematical operations showed positive correlations between the learning process and student learning outcomes in the materials related to arithmetic operations. This study aims to develop an IoT-based mathematical balance tool to support three education 4.0 trends: remote access, personalization, and practice and feedback. This study used modifications of Fahmideh and Zogwhi's IoT development method. There are five phases of IoT development: initialization phase, analysis phase, design phase, implementation phase, and evaluation phase. From each phase of IoT development, IoT-based mathematical balance assistance systems have been successfully built and it complies with the functionality described in the analysis phase. The system performance also shows optimal results with 100% accuracy for reading the student's learning activities. Moreover, it uses less than 10 seconds for processing 1000 data requests.

Author 1: Novian Anggis Suwastika

Author 2: Yovan Julio Adam

Author 3: Rizka Reza Pahlevi

Author 4: Maslin Masrom

Keywords: Arithmetic operation; education 4.0; internet of things development; math balance aids

Paper 27: An Efficient Patient Activity Recognition using LSTM Network and High-Fidelity Body Pose Tracking

Abstract: The need for healthcare services is growing, particularly in light of the COVID-19 epidemic's convoluted trajectory. This causes overcrowding in medical facilities, making it difficult to manage, treat, and monitor patients' health. Therefore, a method to remotely observe the patient's behavior is required, to aid in early warning and treatment, and to reduce the need for hospitalization for patients with minor diseases. This paper proposes a new real-time smart camera system to monitor, recognize and warn the patient's abnormal actions remotely with reasonable cost and easy to deploy in practice. The key benefit of the proposed methods is that patient actions may be detected without the usage of ambient sensors by employing pictures from a regular video camera. It carries out the detection using high-fidelity human body pose tracking with MediaPipe Pose. Then, the Raspberry Pi 4 device and the LSTM network are used for remote monitoring and real-time classification of patient actions. The test dataset is built from reality and reuses the existing datasets. Our system has been evaluated and tested in practice with over 96.84% accuracy, runs at over 30 frames per second, suitable for real-time execution on mobile devices with limited hardware configuration.

Author 1: Thanh-Nghi Doan

Keywords: Human body pose tracking; LSTM; raspberry Pi 4; patient monitoring system

Paper 28: A Covid-19 Positive Case Prediction and People Movement Restriction Classification

Abstract: The world experienced a pandemic that changed people's daily life due to Coronavirus Disease 2019 (covid-19). In Jakarta, the covid-19 cases were discovered on March 18, 2020, and the case increased uncontrollably until the government conducted a movement restriction called pembatasan sosial berskala besar (PSBB). The effectivity of movement restriction was not evaluated in detail. Therefore, we investigated the covid-19 cases in the PSBB period to understand the contribution of movement restriction. Moreover, a prediction model is proposed to computerize the decision of movement restriction. The models are divided into regression and classification models. The regression model is developed to forecast the number of infected cases. At the same time, the classification model is used to identify the best movement restriction type. We utilize data transformation named Principal Component Analysis (PCA) to reduce the number of features. In our case, the best regression method is Multiple Linear Regression (MLP). Then, the best classification method is the Support Vector Machine (SVM). The MLP results are 148.38, 37036.37, and 0.250336 for Mean Absolute Error (MAE), Mean Square Error (MSE), and R2, respectively. In contrast, the SVM achieved an accuracy of 84.81%. Moreover, the prediction system on the website were successfully deployed.

Author 1: I Made Artha Agastya

Keywords: Covid-19; movement restriction; machine learning; positive case; infected prediction

Paper 29: Evaluation of Parameter Fine-Tuning with Transfer Learning for Osteoporosis Classification in Knee Radiograph

Abstract: Osteoporosis is a bone disease that raises the risk of fracture due to the density of the bone mineral being low and the decline of the structure of bone tissue. Among other techniques, such as Dual-Energy X-ray Absorptiometry (DXA), 2D x-ray pictures of the bone can be used to detect osteoporosis. This study aims to evaluate deep convolutional neural networks (CNNs), applied with transfer learning techniques, to categorize specific osteoporosis features in knee radiographs. For objective labeling, we obtained a selection of patient knee x-ray images. The study makes use of the Visual Geometry Group Deep (VGG-16), and VGG-16 with fine-tuning. In this work, the deployed CNNs were assessed using state-of-the-art metrics such as accuracy, sensitivity, and specificity. The evaluation shows that fine-tuning enhanced the VGG-16 CNN's effectiveness for detecting osteoporosis in radiographs of the knee. The accuracy of the VGG-16 with parameter fine-tuning was 88% overall, while the accuracy of the VGG-16 without parameter fine-tuning was 80%.

Author 1: Usman Bello Abubakar

Author 2: Moussa Mahamat Boukar

Author 3: Steve Adeshina

Keywords: Osteoporosis; transfer learning models; convolutional neural network; fine-tuning

Paper 30: Dangerous Goods Container Location Allocation Strategy based on Improved NSGA-II Algorithm

Abstract: The characteristics of port dangerous goods are complicated and diverse in danger, which is very likely to cause chain effects once a fire and explosion accident occurs. Based on the distribution characteristics of dangerous goods container yards and the special national storage requirements for dangerous goods containers, the paper establishes a multi-objective optimization model with a double priority of safety and economy, starting from reducing the number of reversals. The improved non-dominated sorting genetic algorithm based on the elite strategy was used to solve the model and the algorithm was tested and improved. Based on the Pareto optimal solution set, the entropy weight-TOPSIS method was used to optimize the sorting of multiple solution sets, which improved the performance of the algorithm. The analysis further clarifies the important relationship between attributes, and the running time is shortened by 85.7% compared with the traditional NSGA algorithm. The optimization model and algorithm can provide decision support for the actual operation and management of container storage, and provide a good reference for accident risk prevention and control.

Author 1: Xinmei Zhang

Author 2: Nannan Liang

Author 3: Chen Chen

Keywords: Dangerous goods containers; container allocation; improved NSGA-II algorithm; entropy weight-TOPSIS

Paper 31: Word by Word Labelling of Romanized Sindhi Text by using Online Python Tool

Abstract: Sindhi is one of the most ancient languages in the world and it has its own written and spoken scripts. After the rigorous study it was found that a lot of research work has been done in different languages, but word by word labelling of Sindhi language had not been done yet. In this research study, word labelling was done on 100 sentences of Romanized Sindhi texts using Python online tool. The dataset was collected from different sources which include Sindhi newspaper, blogs and social media webpages. From this dataset, a rule-based model has been applied for the Parts-of-Speech (POS) tagging of the Romanized Sindhi sentences. A total of 624 words of Romanized Sindhi texts were tested and successfully tagged by the SindhiNLP tool in which 482 words were tagged as nouns and pronouns, 92 words tagged as verbs and 50 words tagged as determinants.

Author 1: Irum Naz Sodhar

Author 2: Abdul Hafeez Buller

Author 3: Suriani Sulaiman

Author 4: Anam Naz Sodhar

Keywords: Romanized sindhi; word labelling; rule-based model; POS tagging; SindhiNLP tool

Paper 32: Forest Fires Detection using Deep Transfer Learning

Abstract: Forests are vital ecosystems composed of various plant and animal species that have evolved over years to coexist. Such ecosystems are often threatened by wildfires that can start either naturally, as a result of lightning strikes, or unintentionally caused by humans. In general, human-caused fires are more severe and expensive to fight because they are frequently located in inaccessible areas. Wildfires can spread quickly and become extremely dangerous, causing damage to homes and facilities, as well as killing people and animals. Early discovery of wildfires is vital to protect lives, property, and resources. Reinforced imaging technologies can play a key role to detect wildfires earlier. By applying deep learning (DL) over a dataset of images (collected using drones, planes, and satellites), we target to automate the forest fire detection. In this paper, we focus on building a DL model specifically to detect wildfires using transfer learning techniques from the best pretrained DL computer vision architectures available nowadays, such as VGG16, VGG19, Inceptionv3, ResNet50, ResNet50V2, InceptionResNetV2, Xception, Dense-Net, MobileNet, MobileNetV2, and NASNetMobile. Our proposed approach attained a detection rate of more than 99.9% over multiple metrics, proving that it could be used in real-world forest fire detection applications.

Author 1: Mimoun YANDOUZI

Author 2: Mounir GRARI

Author 3: Idriss IDRISSI

Author 4: Mohammed BOUKABOUS

Author 5: Omar MOUSSAOUI

Author 6: Mostafa AZIZI

Author 7: Kamal GHOUMID

Author 8: Aissa KERKOUR ELMIAD

Keywords: Forest fires; wildfires; deep learning; transfer learning; computer vision; convolutional neural networks (CNN)

Paper 33: An Enhancement Technique to Diagnose Colon and Lung Cancer by using Double CLAHE and Deep Learning

Abstract: The most common and deadly cancers are lung and colon cancers. More than a quarter of all cancer cases are caused by them. Early detection of the disease, on the other hand, greatly raises the probability of survival. Image enhancement by Double CLAHE stages and modified neural networks are made to improve classification accuracy and use Deep Learning (DL) algorithms to automate cancer detection. A new Artificial Intelligent classification system is presented in this research to recognize five kinds of colon and lung tissues, three malignant and two benign, with three classes for lung cancer and two classes for colon cancer, based on histological images. The results of the study imply that the suggested system can accurately identify tissues of cancer up to 99.5%. The use of this model will aid medical professionals in the development of an automatic and reliable system for detecting different kinds of colon and lung tumors.

Author 1: Nora yahia Ibrahim

Author 2: Amira Samy Talaat

Keywords: Artificial intelligent system; machine learning; cancer detection; image classification; deep learning

Paper 34: Mobile Applications for the Implementation of Health Control against Covid-19 in Educational Centers, a Systematic Review of the Literature

Abstract: A health crisis caused by the SARS-CoV-2 virus is still ongoing. That is why an important factor for the resumption of on-site classes is the creation of sanitary measures to help control Covid-19. The present research is a literature review, The PRISMA methodology is used and 265 articles are collected from various databases such as EBSCO Host, IEEE Xplore, SAGE, ScienceDirect, and Scopus. According to the inclusion and exclusion criteria, the most relevant articles aligned to the topic were identified, systematizing 119 articles. Showcasing digital technologies used in mobile applications that allow better control, tracking, and monitoring of the health status of students, teachers, and staff of educational centers, in addition to the parameters and quality attributes that must be taken into account for the effective sanitary control of the disease, finally, a development model is proposed.

Author 1: Bryan Quispe-Lavalle

Author 2: Fernando Sierra-Liñan

Author 3: Michael Cabanillas-Carbonell

Keywords: Mobile application; sanitary control; systematic review; digital technologies

Paper 35: Modelling of IoT-WSN Enabled ECG Monitoring System for Patient Queue Updation

Abstract: The advancement of communication technologies has led to the interconnection of different sensors using the Internet of Things (IoT) and Wireless Sensor Network (WSN). WSN for healthcare applications has expanded exponentially due to evolving advantages such as low power requirement of sensors, transmission accuracy, and cost-efficiency. For heart attack patients, the future lies in ECG monitoring in which wearable sensors can be used to acquire patient information. In this paper, an attempt has been made to develop a novel IoT-enabled WSN to record patient information for detection of heart attack and to update queue of patients to ensure prioritized medical attention to critical patients. In the WSN, the Rayleigh Fading channel has been used to transmit data that can be accessed using the cloud repository by the medical staff remotely. The distance from the patient to the medical staff is calculated using Euclidean distance. Further, SNR in comparison to throughput and BER has been computed. The higher SNR indicates the maximum information transfer from patient to hospital staff. The proposed system uses the Grasshopper Optimization and CBNN based disease classification system and bubble sort algorithm has been used for updating patient queue. The proposed GHOA and CBNN has shown improved accuracy of 2.14% over existing techniques like CNN which has accuracy around 82% for R-R feature selection of ECG signals as compared to 82.72% shown by GHOA-CBNN.

Author 1: Parminder Kaur

Author 2: Hardeep Singh Saini

Author 3: Bikrampal Kaur

Keywords: WSN; cloud; ECG monitoring; wearable sensors; IoT; queue updation

Paper 36: A Proposed Deep Learning based Framework for Arabic Text Classification

Abstract: Deep learning has become one of the crucial trends in the modern era due to the huge amount of data that has become available. This paper aims to investigate and improve a generic framework for Arabic Text Classification (ATC) with different deep learning techniques. Besides, it deals directly with a word in its original style as a basic unit of modern Arabic sentence and on a different level of N-grams versus a combination of Intersected Consecutive Word proposed method (ICW). However, it aimed to discuss the results of the different experiments for the enhancements of the proposed method on different deep learning algorithms such as Scaled Conjugate Gradient (SCG) and Gradient descent with momentum and adaptive learning rate backpropagation (GDX) on ATC. The results showed that the proposed framework applied with the SCG algorithm and TF-IDF outperforms the GDX algorithm with an accuracy ratio of 90.65%.

Author 1: Mostafa Sayed

Author 2: Hatem Abdelkader

Author 3: Ayman E. Khedr

Author 4: Rashed Salem

Keywords: Text classification; arabic text classification; scaled conjugate gradient; TF-IDF; GDX; ICW

Paper 37: Simultaneous Importance-Performance Analysis based on SWOT in the Service Domain of Electronic-based Government Systems

Abstract: Decision makers for decades have used SWOT analysis for strategic planning. However, the problems that arise in the SWOT analysis are subjective, so decision-making becomes inefficient. Therefore, SWOT analysis is often combined with other methods to make decision-making strategies more focused and measurable according to priority interests. The SWOT analysis basis in this study is Simultaneous Importance-Performance (SIPA) analysis by observing each indicator's weights. In addition, this study proposes a new method by focusing on competitor factors in strategies mapping to improve services for Electronic-Based Government Systems (SPBE). The object of this study was two local governments in Indonesia, namely the Meranti Islands Regency and the Limapuluh Kota Regency. The results showed that a SIPA-based SWOT analysis has succeeded in showing the Strengths, Weaknesses, Opportunities, and Challenges of the district government. Furthermore, based on the results of hypothesis testing, SIPA-based SWOT identification has reflected a valid organizational situation.

Author 1: Tenia Wahyuningrum

Author 2: Gita Fadila Fitriana

Author 3: Arief Rais Bahtiar

Author 4: Aina Azalea

Author 5: Darwan

Keywords: Importance performance analysis; strength weakness opportunity threat analysis; service quality; electronic based government systems

Paper 38: Federated Learning and its Applications for Security and Communication

Abstract: The not so long ago, Artificial Intelligence (AI) has revolutionized our life by giving rise to the idea of self-learning in different environments. Amongst its different variants, Federated Learning (FL) is a novel approach that relies on decentralized communication data and its associated training. While reducing the amount of data acquired from users, federated learning derives the benefits of popular machine learning techniques, it brings learning to the edge or directly on-device. FL, frequently referred to as a new dawn in AI, is still in its early stages and is yet to garner widespread acceptance, owing to its (unknown) security and privacy implications. In this paper, we give an illustrative explanation of FL techniques, communication, and applications with privacy as well as security issues. According to our findings, there are fewer privacy-specific dangers linked with FL than security threats. We conclude the paper with the challenges of FL with special emphases on security.

Author 1: Hafiz M. Asif

Author 2: Mohamed Abdul Karim

Author 3: Firdous Kausar

Keywords: Federated learning; communication; security; deep learning; Artificial Intelligence

Paper 39: Machine Learning in OCR Technology: Performance Analysis of Different OCR Methods for Slide-to-Text Conversion in Lecture Videos

Abstract: A significant percentage of a lecture video's content shown is text. Video text can therefore be a crucial source for automated video indexing. Researchers have recognised printed and handwritten text extracted from pictures using a variety of machine learning techniques and tools before digitising it. A machine learning technology called optical character recognition (OCR) enables us to recognise and retrieve text information from documents, converting it into searchable and editable data. This study primarily focuses on text extraction from lecture slides using Google Cloud Vision (GCV), Tesseract, Abbyy Finereader, and Transym OCR and compares the results to develop a lecture video indexing scheme for the non-linear steering in lecture videos to watch only the interesting points of topics. We have taken a total of 438 key-frames in 10 categories from seven different lecture videos that range in length. First, binary and greyscale versions of the input colour images are created. Before using the OCR APIs, the frames are additionally preprocessed to improve the image quality. The recognition accuracy demonstrated that the GCV OCR performs effectively, saving computing time by collecting image text with the highest accuracy of other tools, 96.7 percent.

Author 1: Geeta S Hukkeri

Author 2: R H Goudar

Author 3: Prashant Janagond

Author 4: Pooja S Patil

Keywords: Video lectures; keyframes; Google cloud vision (GCV); Tesseract; Abbyy Finereader; Transym; text extraction

Paper 40: Disease Prediction Model based on Neural Network ARIMA Algorithm

Abstract: Because the morbidity data of infectious diseases do not only have a single linear or nonlinear characteristic, but also have both linear and nonlinear characteristics, the combination model prediction method is often used to predict the morbidity of infectious diseases in recent years. Compared with the single model prediction analysis method, the combination model can combine the advantages of a single model to extract the effective information contained in the original time series more scientifically and fully. In the context of big data, for the medical field, massive medical data is complex, and the traditional manual data processing method has been unable to meet the current needs. With the help of the computer, data mining can discover new knowledge that is potentially useful and understandable by clearing, integrating, selecting, and transforming the original data. Using data mining, we can organize and reproduce the useful medical knowledge hidden in medical big data. In this paper, an ARIMA-GRNN model is established; the fitting value and the corresponding time are used as the input of the neural network. The actual morbidity is used as the output to train the network and construct the ARIMA-GRNN combined model. Due to the different information flow of BP neural network and neural network, this study also constructed ARIMA-GRNN combined model and ARIMA model, and compared the modeling effect and prediction performance of various models. The average absolute percentage error of the experimental results in this paper is less than 8.63%, and the average absolute percentage error is less than 5%. Compared with other models, it has a better prediction effect, higher accuracy, and more obvious advantages. In this paper, the prediction of disease is dynamic and continuous. It is of great significance for disease prevention and control to use monitoring data to study the epidemic trend and periodic change law, and to make a reasonable prediction.

Author 1: Kedong Li

Keywords: Disease prevention and control; trend prediction; neural network; combination model; ARIMA algorithm

Paper 41: Evaluation of Spiral Pattern Watermarking Scheme for Common Attacks to Social Media Images

Abstract: The 21st century might be considered the "boom" period for social networking due to the fast expansion of social media use. In terms of user privacy and security regulations, a plethora of new requirements, issues, and concerns have arisen due to the proliferation of social media. With the increase in social media use, images on social media are often modified or fabricated for certain purposes. Therefore, this work implements and evaluates the SPIRAL-LSB algorithm for common attacks for social media images. Image compression was also discussed as images published to social media platforms was often compressed. An analysis was performed to assess the algorithm's output on social media images. The experiments were carried out prior to and after uploading to the Instagram platform. The dataset was subjected to image splicing, copy-move, cut-and-paste, text insertion, and 3D-sticker insertion attacks. The outcome of SPIRAL-LSB was effective for text insertion attacks solely. Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) were selected as the experiment's metrics. The average PSNR value is 63.25, and the SSIM value is 0.99964, both of which are regarded high. This indicates that the watermark has not degraded the quality of the images. This work was designed for usage on social media for intellectual property reasons and may be used to validate the validity of social media images and prevent issues with image integrity, such as image manipulation.

Author 1: Tiew Boon Li

Author 2: Jasni Mohamad Zain

Author 3: Syifak Izhar Hisham

Author 4: Alya Afikah Usop

Keywords: Spiral pattern; fragile watermarking; social media; LSB substitution

Paper 42: Computational Study of Quantum Coherence from Classical Nonlinear Compton Scattering with Strong Fields

Abstract: From the covariant formulation of radiation intensity of Hartemann-Kerman model entirely constructed in the classical electrodynamics scenario, a formulation of coherent states has been obtained in an explicit manner represented by the infinite sum of integer-order Bessel functions. Both linear and nonlinear Compton scattering are included, suggesting that Compton processes can be perceived as coherent states of light-matter interaction.

Author 1: Huber Nieto-Chaupis

Keywords: Quantum coherence; bessel; compton scattering

Paper 43: Cybersecurity Risk Assessment: Modeling Factors Associated with Higher Education Institutions

Abstract: Most universities rely heavily on Information Technology (IT) to process their information and support their vision and mission. This rapid advancement in internet technology leads to increased cyberattacks in Higher Education Institutions (HEIs). To secure their infrastructure from cyberattacks, they must implement the best cybersecurity risk management approach, which involves technological and education-based solutions, to safeguard their environment. However, the main challenges in existing cybersecurity risk management approaches are limited knowledge of how organizations can determine or minimize the significance of risks. As a result, this research seeks to advance understanding to establish a risk assessment model for universities to measure and evaluate the risk in HEIs. The proposed model is based on theoretical aspects that we organized as follows: First, we review the existing cybersecurity frameworks to identify the suitability and limitation of each model. Next, we review current works on cybersecurity risk assessment in HEIs to evaluate the proposed risk assessment approaches, scope and steps. Based on the information gathered, we developed a risk assessment model. Finally, we conclude the study with directions for future research. The result presented from this study may give an insig1ht for HEIs staff to analyze what is to be assessed, how to measure the severity of the risk, and determine the level of risk acceptance, improving their decision-making on risk management.

Author 1: Rachel Ganesen

Author 2: Asmidar Abu Bakar

Author 3: Ramona Ramli

Author 4: Fiza Abdul Rahim

Author 5: Md Nabil Ahmad Zawawi

Keywords: Cyber security; risk assessment; university

Paper 44: Acne Classification with Gaussian Mixture Model based on Texture Features

Abstract: This paper presents an acne detection method on face images using a Gaussian Mixture Model (GMM). First, the skin area in the face image is segmented based on color information using the GMM. Second, the candidates of the acne region are then extracted using a Laplacian of Gaussian-based blob detection strategy. Then, texture features are extracted from acne candidates using either a Gabor Filter or Gray Level Co-occurrence Matrix (GLCM). Lastly, these features are then utilized as input in the GMM for verifying whether these regions are acne or not. In our experiment, the proposed method was evaluated using face images from ACNE04 dataset. Based on the experiment, it is found that the best classification results were obtained when GLCM features in the Cr-YCbCr channel are applied. In addition, the proposed method has competitive performance compared to K-Nearest Neighbor (KNN).

Author 1: Alfa Nadhya Maimanah

Author 2: Wahyono

Author 3: Faizal Makhrus

Keywords: Acne; GLCM; Gabor filter; Gaussian mixture model

Paper 45: Learning Content Classification and Mapping Content to Synonymous Learners based on 2022 Augmented Verb List of Marzano and Kendall Taxonomy

Abstract: Finding suitable learning content for learners with different learning styles is a challenging task in the learning process. Hence it is essential to follow some learning taxonomies to deliver learner-centric learner content. Learning taxonomies are used to express various learning practices and learning habits to be followed by the learner for a better learning process. The investigator has already classified the learners based on the 2022 augmented verb list of Marzano and Kendall taxonomy. The main objective of this paper is to minutely classify the tutor-defined learning contents according to the domains as well as the subdomains of the considered taxonomy which is in text format. Providing personalized learning content could help the learners for a better understanding of learning content and their interrelationship which in turn produce better learning outcomes. Mapping the six levels of learning contents into the corresponding learner is a challenging task. Hence the investigator has chosen seven algorithms including Bagging, XG Boost, Support Vector Machine from Machine Learning and four algorithms including Convolutional Neural Network, and Deep Neural Network in Deep Learning algorithm to classify the learning contents. The experimental results indicate that Support Vector Machine performed well in machine learning and Deep Neural Network yields good performance in deep learning in the learning content classification process. These micro contents were organized using a property graph. Further, the micro contents were retrieved from the property graph using SPARQL for mapping the classified contents to the corresponding learners to achieve personalization in the learning process.

Author 1: S. Celine

Author 2: M. Maria Dominic

Author 3: F. Sagayaraj Fransis

Author 4: M. Savitha Devi

Keywords: Learning taxonomies; marzano and kendall taxonomy; personalization; XG boost; deep neural network; CNN; property graph; action verbs; content classification

Paper 46: The Hybrid Combinatorial Design-based Session Key Distribution Method for IoT Networks

Abstract: Internet of Things (IoT) is currently being used in a range of applications as cutting-edge technology. IoT is a technological platform that connects the physical and digital worlds, allowing us to use things remotely. Various sensor-connected nodes serve as objects that communicate with one another over the internet. Hence security-related problems are more likely to arise in IoT networks. However, due to resource constraints such as power and memory capacity, complex security algorithms cannot be implemented in IoT networks. One of the security measures for IoT networks is to implement the lightweight key distribution algorithm. The lightweight key management process is essential for IoT networks to share the key securely. We presented the new key-distribution approach based on the hybrid combinatorial design that implements lightweight algorithms and describes the analysis functions. The comparison to existing hybrid combinatorial works shows better connectivity, resilience, and scalability.

Author 1: Gundala Venkata Hindumathi

Author 2: D. Lalitha Bhaskari

Keywords: Key distribution; hybrid combinatorial design; IoT networks; resource constraint nodes; symmetric key generation

Paper 47: Automated Study Plan Generator using Rule-based and Knapsack Problem

Abstract: Undergraduate students are given the flexibility of arranging courses throughout their study duration especially when they are eligible for credit exemption for the courses taken during their diploma study. Issues arise when students arrange their studies manually. Improper course arrangement in the study plan may be resulting some of the selected courses do not correspond to the courses offered, and imbalance credit hours. Hence, this study aims to propose an algorithm to generate an automated and accurate study plan throughout the study duration. A combination of rule-based and knapsack problem were proposed to generate an automated study plan. A quantitative methodology through expert’s reviews and questionnaire survey was conducted to evaluate the accuracy of the proposed algorithm. The proposed algorithm shows high accuracy. In conclusion, the combination of rule-based and knapsack problem is appropriate to generate an automated and accurate study plan. The automated study plan generator can help students generate an effective study plan.

Author 1: Muhammad Amin Mustapa

Author 2: Lizawati Salahuddin

Author 3: Ummi Rabaah Hashim

Keywords: Knapsack problem; rule-based; study plan; undergraduate; credit exemption

Paper 48: Combining Multiple Classifiers using Ensemble Method for Anomaly Detection in Blockchain Networks: A Comprehensive Review

Abstract: Blockchain is one of the most anticipated technology revolutions, with immense promise in various applications. It is a distributed and encrypted database that can address a range of challenges connected to online security and trust. While many people identify Blockchain with cryptocurrencies such as Bitcoin, it has a wide range of applications in supply chain management, health, Internet of Things (IoT), education, identity theft prevention, logistics, and the execution of digital smart contracts. Although Blockchain Technology (BT) has numerous advantages for Decentralized Applications (DApps), it is nevertheless vulnerable to abuse, smart contract failures, security, theft, trespassing, and other concerns. As a result, using Machine Learning (ML) models to detect anomalies is an excellent way to detect and safeguard blockchain networks from criminal activity. Adapting ensemble learning methods in ML to create better prediction outcomes is a viable approach for anomaly identification. Ensemble learning, as the name implies, refers to creating a stronger and more accurate classification by combining the prediction results of numerous weak models. As a result, an in-depth evaluation of ensemble learning methodologies for anomaly detection in the blockchain network ecosystem is applied in this paper. It comprises numerous ensemble methods (e.g., averaging, voting, stacking, boosting, bagging). The review collects data from three established databases, which are Scopus, Web of Science (WoS), and Google Scholar. Specific keywords are employed, such as Blockchain, Ethereum, Bitcoin, Anomaly Detection, and Ensemble Learning, employing advanced searching algorithms. The results of the search found 60 primary articles from 2017 to 2022 (30 from Scopus, 20 from the WoS, and 10 from Google Scholar). Based on these findings, we decided to divide our debate into three primary themes: (1) the fundamentals of Blockchain Technology (BT), (2) the overview of ensemble learning, and (3) the integration and analysis of ensemble learning in blockchain networks for anomaly detection. In terms of awareness and knowledge, the results are also discussed in terms of what they mean and where future research should go.

Author 1: Sabri Hisham

Author 2: Mokhairi Makhtar

Author 3: Azwa Abdul Aziz

Keywords: Blockchain; Ethereum; Bitcoin; ensemble; anomaly detection

Paper 49: A Novel Hybrid Sentiment Analysis Classification Approach for Mobile Applications Arabic Slang Reviews

Abstract: Arabic language incurs from the shortage of accessible huge datasets for Sentiment Analysis (SA), Machine Learning (ML), and Deep Learning (DL) applications. In this paper, we present MASR, a simple Mobile Applications Arabic Slang Reviews dataset for SA, ML, and DL applications which comprises of 2469 Egyptian Mobile Apps reviews, and help app developers meet user requirements evolution. Our methodology consists of six phases. We collect mobile apps reviews dataset, then apply preprocessing steps, in addition perform SA tasks. To evaluate MASR datasets, first we apply ML classification techniques: K-Nearest Neighbors (K-NN), Support vector machine (SVM), Logistic Regression (LR), and Random Forest (RF), and DL classification technique: Multi-layer Perceptron Neural Network (MLP-NN). From the examination for pervious classification techniques, we adopted a hybrid classification approach combined from the top two ML classifier accuracy results (LR, RF), and DL classifier (MLP-NN). The findings prove the adequacy of a hybrid supervised classification approach for MASR datasets.

Author 1: Rabab Emad Saudy

Author 2: Alaa El Din M. El-Ghazaly

Author 3: Eman S. Nasr

Author 4: Mervat H. Gheith

Keywords: Arabic sentiment analysis; mobile application; hybrid classification model; hybrid supervised classification approach; Google play store; random forest; logistic regression; neural network; multi-layer perceptron neural network; machine learning; deep learning

Paper 50: User Evaluation of UbiQuitous Access Learning (UQAL) Portal: Measuring User Experience

Abstract: The goal of user experience (UX) research in human-computer interaction is to understand how humans interact with technology. This paper aimed to evaluate the interface and user experience of UbiQuitous Access Learning Portal (UQAL) and make recommendations for the system interface. UQAL Portal is an e-learning web portal that teaches a targeted group of users how to start a business or an online business using an e-learning portal. The portal will be used to search for business-related information, among other things. The User Experience Questionnaire (UEQ) is used to evaluate user experience. The interface is evaluated using a heuristic evaluation technique based on Nielsen’s ten heuristics. According to the UEQ results, the average score for each aspect in 30 UQAL users is: Attractiveness aspect: 1.77; Perspicuity aspect: 2.20; Efficiency aspect: 2.30; Dependability aspect: 1.73; Stimulation aspect: 0.63; and Novelty aspect: 1.27. A comparison of the average score in the dataset product of UEQ Data Analysis Tool revealed that the Perspicuity, Efficiency, and Dependability aspects of UQAL belonged to the Excellent category. The Attractiveness and Novelty aspects could be categorized as Good, and its stimulation could be categorized as Below Average. Four evaluators participate in the heuristic evaluation, which tests all user categories in UQAL. The findings of this study can be used as a suggestion and reference for UQAL Portal improvement.

Author 1: Nazlena Mohamad Ali

Author 2: Wan Fatimah Wan Ahmad

Author 3: Zainab Abu Bakar

Keywords: User experience questionnaire; user experience; user interface; heuristic evaluation

Paper 51: Design of a Cloud-Blockchain-based Secure Internet of Things Architecture

Abstract: The growing number of Internet of Things (IoT) objects and the operational and security challenges in IoT systems are encouraging researchers to design suitable IoT architecture. Enormous data generated in the IoT environment face several kinds of security and privacy challenges. IoT system generally suffers from several issues like data storage, safety, privacy, integrity, transparency, trust, and single point of failure. IoT environment is emerging with several solutions to resolve these problems. The main objective of this paper is to design a cloud-blockchain-based secure IoT architecture that provides advanced and efficient storage and security solutions to IoT ecosystem. Blockchain technology appears to be a decent choice to resolve such kinds of problems. Blockchain technology uses a hash-based cryptographic technique for information security and integrity. Cloud computing provides advanced storage solutions with several remote services to store, compute and analyze the data. The proposed IoT architecture is based on the integration of cloud and blockchain services, which aim to provide transparent, decentralized, and trustworthy and secure storage solutions. In addition to the standard layers (perception layer, network layer, processing layer, and application layer) the proposed IoT architecture in the present paper includes a service layer, a security layer, and a parallel management and control layer, which focus on the security and management of the entire IoT infrastructure.

Author 1: Deepti Rani

Author 2: Nasib Singh Gill

Author 3: Preeti Gulia

Keywords: Internet of things; cloud computing; blockchain; iot architecture; security and services

Paper 52: Medical Big Data Analysis using Binary Moth-Flame with Whale Optimization Approach

Abstract: The accurate analysis of medical data is dependent on early disease detection and the value of accuracy is reduced when the medical data quality is poor. However, existing techniques have lower efficiency in handling heterogeneous medical data and the complexity of the features was not enhanced using an optimal feature selection model. The present research work has used the machine learning algorithm effectively for chronic disease prediction such as heart disease, cancer, diabetes, stroke, and arthritis for the frequent communities. The detailed information about the attributes is required to be known as it is significant in analyzing the medical data. The process of selecting the attributes plays an important role in decision-making for medical disease analysis. This research proposes Binary Moth-Flame Optimization (B-MFO) for effective feature selection to achieve higher performance in small and medium datasets. Additionally, the Whale Optimization Algorithm (WOA) is used that showed better performances for LSTM that suited well for the process of classification to predict the time series. The present research work utilizes Spark Streaming layers for data streaming to diagnose using Long Short Term Memory (LSTM) with whale optimization approach which is from the heterogeneous medical data. The proposed B-MFO-WOA method results showed that the proposed method obtained 97.45% accuracy better compared to the existing Modified adaptive neuro fuzzy inference system of 95.91% of accuracy and B-MFO of 92.43 % accuracy for the models.

Author 1: Saka Uma Maheswara Rao

Author 2: K Venkata Rao

Author 3: Prasad Reddy PVGD

Keywords: Binary moth-flame optimization; complexity of the features; medical data; long short term memory; spark streaming layers; whale optimization algorithm

Paper 53: Implementation of a Mobile Application based on the Convolutional Neural Network for the Diagnosis of Pneumonia

Abstract: Pneumonia is the main cause of infant mortality in Peru, which has led to plansfig, such as vaccination campaigns, greater economic investment in health, and the strengthening of specialized medical personnel, however, mortality rates remain high. In this sense, the implementation of new computer technologies such as Deep Learning through the use of the artificial neural network is proposed. The objective of this project was to determine the influence of a mobile application based on a Convolutional Neural Network for the diagnosis of Pneumonia, the project consists of the analysis of images of Chest X-rays with Pneumonia and Normal by means of an application developed called “Diagnost”. The study was carried out considering a control group and a study group formed by 33 medical staff members who used the application. The analysis of the data obtained was made based on the study of 3 indicators, detection time, result in accuracy, and reduction of medical assistance. According to the results, it was concluded that the mobile application based on the convolutional neural network allows the early detection of Pneumonia and allows the reduction of medical assistance, however, it is still necessary to continue working on the accuracy of the diagnosis.

Author 1: Jazmin Flores-Rodriguez

Author 2: Michael Cabanillas-Carbonell

Keywords: Pneumonia; convolutional neural network; deep learning; chest x-rays

Paper 54: Parameter Estimation in Computational Systems Biology Models: A Comparative Study of Initialization Methods in Global Optimization

Abstract: This paper compares different initialization methods and investigates their performance and effects on estimating kinetic parameters’ value in models of biological systems. Estimating parameters values is difficult and time-consuming process due to their highly nonlinear and huge number of kinetic parameters involved. Global optimization method based on an enhanced scatter search (ESS) algorithm is a suitable choice to address this issue. However, despite its resounding success, the performance of ESS may decrease in solving high dimension problem. In this work, several choices of initialization methods are compared and experimental results indicated that the algorithm is sensitive to the initial value of kinetic parameters. Statistical results revealed that uniformly distributed random number generator (RNG) and controlled randomization (CR) that being used in ESS may lead to poor algorithm performance. In addition, the different initialization methods also influenced model accuracy. Our proposed methodology shows that initialization based on opposition-based learning scheme have shown 10% better accuracy in term of cost function.

Author 1: Muhammad Akmal Remli

Author 2: Nor-Syahidatul N.Ismail

Author 3: Noor Azida Sahabudin

Author 4: Nor Bakiah Abd Warif

Keywords: Metaheuristic; opposition-based learning; kinetic parameters; initialization method; metabolic engineering

Paper 55: Determinants of Information Security Awareness and Behaviour Strategies in Public Sector Organizations among Employees

Abstract: In this digital era, protecting an organisation's sensitive information system assets against cyberattacks is challenging. Globally, organisations spend heavily on information security (InfoSec) technological countermeasures. Public and private sectors often fail to secure their information assets because they depend primarily on technical solutions. Human components create the bulk of cybersecurity incidents directly or indirectly, causing many organisational information security breaches. Employees' information security awareness (ISA) is crucial to preventing poor information security behaviours. Until recently, there was little combined information on how to improve ISA and how investigated factors influencing employees' ISA levels were. This paper proposed a comprehensive theoretical model based on the Protection Motivation Theory, the Theory of Planned Behaviour, the General Deterrence Theory, and Facilitating Conditions for assessing public sector employees' ISA intentions for information security behaviour. Using a survey and the structural equation modelling (SEM) method, this research reveals that the utilised factors are positively associated with actual information security behaviour adoption, except for perceived sanction certainty. The findings suggest that the three theories and facilitating conditions provide the most influential theoretical framework for explaining public sector employees' information security adoption behaviour. These findings support previous empirical research on why employees' information on security behaviours vary. Consistent with earlier research, these psychological factors are just as critical as facilitating conditions in ensuring more significant behavioural intention to engage in ISA activities, ensuring information security behaviour. The study recommends that public-sector organisations invest in their employees' applied information security training.

Author 1: Al-Shanfari I

Author 2: Warusia Yassin

Author 3: Nasser Tabook

Author 4: Roesnita Ismail

Author 5: Anuar Ismail

Keywords: Information security awareness; behaviour strategies; self-administered questionnaire; structural equation modelling (SEM)

Paper 56: Novel Oversampling Algorithm for Handling Imbalanced Data Classification

Abstract: In the current age, the attention of researchers is immersed by numerous imbalanced data applications. These application areas are intrusion detection in security, fraud recognition in finance, medical applications dealing with disease diagnosis pilfering in electricity, and many more. Imbalanced data applications are categorized into two types: binary and multiclass data imbalance. Unequal data distribution among data diverts classification performance metrics towards the majority data instance class and ignores the minority data, instance class. Data imbalance leads to an increase in the classification error rate. Random Forest Classification (RFC) is best suitable technique to deal with imbalanced datasets. This paper proposes the novel oversampling rate calculation algorithm as Improvised Dynamic Binary-Multiclass Imbalanced Oversampling Rate (IDBMORate). Experimentation analysis of the proposed novel approach IDBMORate on Page-block (Binary) dataset shows that instances of positive class is increased from 559 to 1118 whereas negative instance class remains same as 4913. In case of referred multiclass dataset (Ecoli), IDBMORate produces the consistent result as minority classes (om, omL, imS, imL) instances are oversampled majority class instances remains unchanged. IDBMORate algorithm reduces the ignorance of minority class and oversamples its data without disturbing the size of the majority instance class. Thus, it reduces the overall computation cost and leads towards the improvisation of classification performance.

Author 1: Anjali S. More

Author 2: Dipti P. Rana

Keywords: Binary imbalance; multiclass imbalance; oversampling; random forest classification; classification

Paper 57: Predicting Malicious Software in IoT Environment Based on Machine Learning and Data Mining Techniques

Abstract: The Internet of Things (IoT) enable the IoT to sense and respond using the power of computing to autonomously come up with the best solutions for any industry today. However, Internet of Things have vulnerabilities since it can be hacked by cybercriminals. The cybercriminals know where the IoT vulnerabilities are, such as unsecured update mechanisms and malware (Malicious Software) to attack the IoT devices. The recently posted IoT-23 dataset based on several IoT devices such as Philips Hue, Amazon Echo devices and Somfy door lock were used for machine learning classification algorithms and data mining techniques with training and testing for predictive modelling of a variety of malware attacks like Distributed Denial of Service (DDoS), Command and Control (C&C) and various IoT botnets like Mirai and Okiru. This paper aims to develop predictive modeling that will predict malicious software to protect IoT and reduce vulnerabilities by using machine learning and data mining techniques. We collected, analyzed and processed benign and several of malicious software in IoT network traffic. Malware prediction is crucial in maintaining IoT devices’ safety and security from cybercriminals’ activities. Furthermore, the Principal Component Analysis (PCA) method was applied to determine the important features of IoT-23. In addition, this study compared with previous studies that used the IoT-23 dataset in terms of accuracy rate and other metrics. Experiments show that Random Forest (RF) classifier achieved the predictive model produced classification accuracy 0.9714% as well as predict 8754 samples with various types of malware and obtained 0.9644% of Area Under Curve (AUC) which outperforms several bassline machine learning classification models.

Author 1: Abdulmohsen Alharbi

Author 2: Md. Abdul Hamid

Author 3: Husam Lahza

Keywords: Machine learning; internet of things; malware; predictive modeling; cyber threats

Paper 58: Power user Data Feature Matching Verification Model based on TSVM Semi-supervised Learning Algorithm

Abstract: The existing model for identifying user data features based on smart meter data adopts a supervised learning method. Although the model has good identification performance under the condition of sufficient index samples, matching data are difficult to obtain and the marking cost is high in real life. The identification accuracy is significantly reduced when the matching data are insufficient or unavailable in the supervised learning method. In view of the above problems, based on the smart meter data, this paper proposes a feature recognition method for residential user data based on semi-supervised learning, which uses three indicators to evaluate the recognition performance of the proposed semi-supervised learning method for residential user data features and to find the appropriate feature selection method and data acquisition resolution. Then, explore the role of this method in real life when there is insufficient or unavailable matching data. Experimental results show that the performance of the proposed semi-supervised learning algorithm is better than that of the supervised learning algorithm, and the accuracy of the proposed algorithm is better than or close to that of the supervised learning algorithm.

Author 1: Yakui Zhu

Author 2: Rui Zhang

Author 3: Xiaoxiao Lu

Keywords: Power system; data matching; data characteristics; semi-supervised learning algorithm; load model

Paper 59: Light Gradient Boosting with Hyper Parameter Tuning Optimization for COVID-19 Prediction

Abstract: The 2019 coronavirus disease (COVID-19) caused pandemic and a huge number of deaths in the world. COVID-19 screening is needed to identify suspected positive COVID-19 or not and it can reduce the spread of COVID-19. The polymerase chain reaction (PCR) test for COVID-19 is a test that analyzes the respiratory specimen. The blood test also can be used to show people who have been infected with SARS-CoV-2. In addition, age parameters also contribute to the susceptibility of COVID-19 transmission. This paper presents the extra trees classification with random over-sampling by considering blood and age parameters for COVID-19 screening. This research proposes enhanced preprocessing data by using KNN Imputer to handle large missing values. The experiments evaluated the existing classification methods such as Random Forest, Extra Trees, Ada Boost, Gradient Boosting, and the proposed Light Gradient Boosting with hyperparameter tuning to measure the predictions of patients infected with SARS-CoV-2. The experiments used Albert Einstein Hospital test data in Brazil that consisted of 5,644 sample data from 559 patients with infected SARS-CoV-2. The experimental results show that the proposed scheme achieves an accuracy of about 98,58%, recall of 98,58%, the precision of 98,61%, F1-Score of 98,61%, and AUC of 0,9682.

Author 1: Ferda Ernawan

Author 2: Kartika Handayani

Author 3: Mohammad Fakhreldin

Author 4: Yagoub Abbker

Keywords: ROS; light gradient boosting; hyper parameter tuning; COVID-19 screening; blood and age based

Paper 60: Surface Electromyography Signal Classification for the Detection of Temporomandibular Joint Disorder using Spectral Mapping Method

Abstract: Temporomandibular joint Disorder (TMD) is with multifaceted and complex signs and symptoms which makes day to day activities of an individual uneasy. Electromyographic (EMG) processing of related muscles recordings could provide an early and immediate detection of TMD. To detect the TMD using surface electromyography (SEMG) of Masseter and Temporalis muscle with discrete wavelet transform (DWT) using spectral coding. To analyze the data, a new feature selection approach in the spectral domain is proposed. For statistical analyses, SPSS version 24 is employed. The results of the study revealed that the proposed approach was able to improve the accuracy of the classification by implementing a combination of DWT and the Support Vector Machine (SVM). The proposed method also exhibited a significant improvement in its performance in terms of its accuracy with 93%. In addition, the statistical analysis revealed that the model was able to improve the mean rank of the experimental and control group.

Author 1: Bormane D. S

Author 2: Roopa B. Kakkeri

Author 3: R. B. Kakkeri

Keywords: Temporomandibular joint (TMJ); temporomandibular joint disorder (TMD); surface electromyography (sEMG); spectral mapping; discrete wavelet transform (DWT)

Paper 61: A Deep Learning Approach for Viral DNA Sequence Classification using Genetic Algorithm

Abstract: DNA sequence classification is one of the major challenges in biological data processing. The identification and classification of novel viral genome sequences drastically help in reducing the dangers of a viral outbreak like COVID-19. The more accurate the classification of these viruses, the faster a vaccine can be produced to counter them. Thus, more accurate methods should be utilized to classify the viral DNA. This research proposes a hybrid deep learning model for efficient viral DNA sequence classification. A genetic algorithm (GA) was utilized for weight optimization with Convolutional Neural Networks (CNN) architecture. Furthermore, Long Short-Term Memory (LSTM) as well as Bidirectional CNN-LSTM model architectures are employed. Encoding methods are needed to transform the DNA into numeric format for the proposed model. Three different encoding methods to represent DNA sequences as input to the proposed model were experimented: k-mer, label encoding, and one hot vector encoding. Furthermore, an efficient oversampling method was applied to overcome the imbalanced dataset issues. The performance of the proposed GA optimized CNN hybrid model using label encoding achieved the highest classification accuracy of 94.88% compared with other encoding methods.

Author 1: Ahmed El-Tohamy

Author 2: Huda Amin Maghwary

Author 3: Nagwa Badr

Keywords: Deep learning; sequence classification; convolutional neural networks; genetic algorithm; sequence encoding

Paper 62: J-Selaras: New Algorithm for Automated Data Integration Tools

Abstract: Data integration is a popular technique or method today for data converting and sharing within new application with different database format and location. The interaction of data from one application system to another application system requires an intermediary software or middleware that allows the data to be transferred or read systematically and easily. The development of dynamic algorithms allows data in various formats, whether structured or unstructured, to be transferred to various types of databases smoothly. A case study was conducted for the Bill of Quantity (BQ) data in the known Excel format generated through CostX software in a single sheet Excel file. It was transferred to a single workbook with multiple sheets with formulation generated automatically. Thus, an algorithm was developed and tested through the development of the J-Selaras System. This algorithm can remove the noisy data or data symbols that are not related in the excel single sheet (CostX) file and automatically transfer to multiple excel sheets with macros formulation quickly. The implementation results indicate a significant contribution where it reduces in execution time of BQ processes and manpower resources used.

Author 1: Mustafa Man

Author 2: Wan Aezwani Wan Abu Bakar

Author 3: Mohd. Kamir Yusof

Author 4: Norisah Abdul Ghani

Author 5: Mohd Adza Arshad

Author 6: Raja Normawati Raja Ayob

Author 7: Kamarul Azhar Mahmood

Author 8: Faizul Azwan Ariffin

Author 9: Mohamad Dizi Che Kadir

Author 10: Lily Mariya Rosli

Author 11: Nurhafiza Binti Abu Yaziz

Keywords: Automated data integration algorithm; bill of quantity (BQ); CostX; single and multiple sheets; J-Selaras tool

Paper 63: Efficient Function Integration and a Case Study with Gompertz Functions for Covid-19 Waves

Abstract: Numerical algorithms are widely used in different applications, therefore, the execution time of the functions involved in numerical algorithms is important, and, in some cases, decisive, for example, in machine learning algorithms. Given a finite set of independent functions A(x), B(x), ..., Z(x) with domains defined by disjoint, consecutive, and not necessarily adjacent intervals, the main goal is to integrate into a single function F(x) = k1×A(x) + k2×B(x) + … + kn×Z(x), where each activation coefficient k, is one if x is in the interval of the respective domain and zero otherwise. The novelty of this work is the presentation and formal demonstration of two general forms of integration of functions in a single function: The first is the mathematical version and the second is the computational version (with the AND function at the bit level), which is characterized by its efficiency. The result is applied in a case study (Peru), where two regression functions were obtained that integrate all the waves of Covid-19, that is, the epidemic curve of the variable global number of deaths/infected per day, the adjustment provided a highly statistically significant measure of correlation, a Pearson's product-moment correlation of 0.96 and 0.98 respectively. Finally, the size of the epidemic was projected for the next 30 days.

Author 1: Oliver Amadeo Vilca-Huayta

Author 2: Ubaldo Yancachajlla Tito

Keywords: Covid-19; corona virus; function integration; Gompertz model; machine learning

Paper 64: Enhancement of Low-Light Image using Homomorphic Filtering, Unsharp Masking, and Gamma Correction

Abstract: Now-a-days, a digital image can be found almost everywhere, and digital image processing plays a huge role in analyzing and enhancing the image so that it can be delivered in a good condition. Color distortion and loss of image details are the common problems that were faced by low-light image enhancement methods. This paper introduces a low-light image enhancement method that applied the concept of homomorphic filtering, unsharp masking, and gamma correction. The aim of the proposed method is to minimize the two problems stated while producing images of better quality when compared to the other low-light image enhancement methods. An objective evaluation was done on the proposed method, comparing the results produced by the enhanced method with other two existing low-light image enhancement methods. The results obtained showed the proposed method outshines the other two existing low-light image enhancement method in maintaining the image details and producing a natural looking image, achieving the lowest Mean Square Error (MSE) and Lightness Order Error (LOE) scores, and has the highest Features Similarity Index color (FSIMc), Features Similarity Index (FSIM), Structure Similarity Index (SSIM), and Visual information fidelity (VIF) scores. Future studies that should be made on this research are to implement dehaze and denoise functionality into the low-light image as well as enabling it to be applicable in real-time scenarios.

Author 1: Tan Wan Yin

Author 2: Kasthuri A/P Subaramaniam

Author 3: Abdul Samad Bin Shibghatullah

Author 4: Nur Farraliza Mansor

Keywords: Low-light image; gamma correction; homomorphic filtering. low-light enhancement; unsharp masking

Paper 65: An Ensemble of Arabic Transformer-based Models for Arabic Sentiment Analysis

Abstract: In recent years, sentiment analysis has gained momentum as a research area. This task aims at identifying the opinion that is expressed in a subjective statement. An opinion is a subjective expression describing personal thoughts and feelings. These thoughts and feelings can be assigned with a certain sentiment. The most studied sentiments are positive, negative, and neutral. Since the introduction of attention mechanism in machine learning, sentiment analysis techniques have evolved from recurrent neural networks to transformer models. Transformer-based models are encoder-decoder systems with attention. Attention mechanism has permitted models to consider only relevant parts of a given sequence. Making use of this feature in encoder-decoder architecture has impacted the performance of transformer models in several natural language processing tasks, including sentiment analysis. A significant number of Arabic transformer-based models have been pre-trained recently to perform Arabic sentiment analysis tasks. Most of these models are implemented based on Bidirectional Encoder Representations from Transformers (BERT) such as AraBERT, CAMeLBERT, Arabic ALBERT and GigaBERT. Recent studies have confirmed the effectiveness of this type of models in Arabic sentiment analysis. Thus, in this work, two transformer-based models, namely AraBERT and CAMeLBERT have been experimented. Furthermore, an ensemble model has been implemented to achieve more reasonable performance.

Author 1: Ikram El Karfi

Author 2: Sanaa El Fkihi

Keywords: Transformers; BERT; ensemble learning; Arabic sentiment analysis

Paper 66: Grover’s Algorithm for Data Lake Optimization Queries

Abstract: Now-a-days, the use of No-SQL databases is one of the potential options for storing and processing big data lakes. However, searching for large data in No-SQL databases is a complex and time-consuming task. Further, information retrieval from big data management suffers in terms of execution time. To reduce the execution time during the search process, we propose a fast and suitable approach based on the quantum Grover algorithm, which represents one of the best-known approaches for searching in an unstructured database and resolves the unsorted search query in O (√n) time complexity. To assess our proposal, a comparative study with linear and binary search algorithms was conducted to prove the effectiveness of Grover's algorithms. Then, we perform extensive experiment evaluations based on ibm_qasm_simulator for searching one item out of eight using Grover’s search algorithm based on three qubits. The experiments outcomes revealed encouraging results, with an accuracy of 0.948, well in accordance with the theoretical result. Moreover, a discussion of the sensitivity of Grover's algorithm through different iterations was carried out. Then, exceeding the optimal number of iterations round (π/4 √N), induces low accuracy of the marked state. Furthermore, the incorrect selection of this parameter can outline the solution.

Author 1: Mohamed CHERRADI

Author 2: Anass EL HADDADI

Keywords: Big data; data management; information retrieval; quantum computing

Paper 67: Inclusive Study of Fake News Detection for COVID-19 with New Dataset using Supervised Learning Algorithms

Abstract: Covid-19 imposes many bans and restrictions on news, individuals and teams, and thus social networks have become one of the most used platforms for sharing and destroying news, which can be either fake or true. Therefore, detecting fake news has become imperative and thus has drawn the attention of researchers to develop approaches for understanding and classifying news content. The focus was on the Twitter platform because it is one of the most used platforms for sharing and disseminating information among many organizations, personalities, news agencies, and satellite stations. In this research, we attempt to improve the detection process of fake news by employing supervised machine learning techniques on our newly developed dataset. Specifically, the proposed system categorizes fake news related to COVID-19 extracted from the Twitter platform using four machine learning-based models, including decision tree (DT), Naïve Bayes (NB), artificial neural network (ANN), and k-nearest neighbors (KNN) classifiers. Besides, the developed detection models were evaluated on our new dataset, which we extracted from Twitter in a real-time process using standard evaluation metrics such as detection accuracy (ACC), F1-score (FSC), the under the curve (AUC), and Matthew's correlation coefficient (MCC). In the first set of experiments which employ the full dataset (i.e., 14,000 tweets), our experimental evaluation reported that DT based detection model had achieved the highest detection performance scoring 99.0%, 96.0%, 98.0%, and 90.0% in ACC, FSC, AUC, and MCC, respectively. The second set of experiments employs the small dataset (i.e., 700 tweets); our experimental evaluation reported that DT based detection model had achieved the highest detection performance scoring 89.5%, 89.5%, 93.0%, and 80.0% in ACC, FSC, AUC, and MCC, respectively. The results obtained for all experiments have been generated for the best-selected features.

Author 1: Emad K. Qalaja

Author 2: Qasem Abu Al-Haija

Author 3: Afaf Tareef

Author 4: Mohammad M. Al-Nabhan

Keywords: Machine learning; fake news; twitter; covid-19; correlation coefficient

Paper 68: A 4-Layered Plan-Driven Model (4LPdM) to Improve Software Development

Abstract: Quality is the degree of excellence of a product and one of the most important factors of software projects that mainly defines user satisfaction and success of the project. Software methodologies represent a variety of tasks, processes, and roles to manage time, cost, and quality. The invention, innovation, and diffusion for technological advancement creates challenges of software projects, thus several existing methodologies albeit with limited scope. A software product is highly influenced by the latest technology and distributed project management opportunities. Management issues are introduced for a virtual project management environment when resource persons are in another corner of the world. To resolve the problem, this research presents a new software project management model (4-LPdM) with alternative actions and practices to effectively manage. The model was presented to 20 different organizations and 29 respondents gave feedback who had experience between 1-16 years in multiple sections of software engineering. The model is evaluated based on the factors of advanced PMBOK 4.0 (scope, cost, quality, resource, risk, plan) and two (management, sustainability) additional features according to the demand of experts. This research illustrates statistical analyses to examine the significance of the proposed model besides a comprehensive comparative study of the traditional methodology.

Author 1: Kamal Uddin Sarker

Author 2: Aziz Bin Deraman

Author 3: Raza Hasan

Author 4: Ali Abbas

Keywords: Software development methodology; project management; 4-layered plan-driven model; quality factors; sustainability

Paper 69: A New Model to Detect COVID-19 Coughing and Breathing Sound Symptoms Classification from CQT and Mel Spectrogram Image Representation using Deep Learning

Abstract: Deep Learning is a relatively new Artificial Intelligence technique that has shown to be extremely effective in a variety of fields. Image categorization and also the identification of artefacts in images are being employed in visual recognition. The goal of this study is to recognize COVID-19 artefacts like cough and also breath noises in signals from real-world situations. The suggested strategy considers two major steps. The first step is a signal-to-image translation that is aided by the Constant-Q Transform (CQT) and a Mel-scale spectrogram method. Next, nine deep transfer models (GoogleNet, ResNet18/34/50/100/101, SqueezeNet, MobileNetv2, and NasNetmobile) are used to extract and also categorise features. The digital audio signal will be represented by the recorded voice. The CQT will transform a time-domain audio input to a frequency-domain signal. To produce a spectrogram, the frequency will really be converted to a log scale as well as the colour dimension will be converted to decibels. To construct a Mel spectrogram, the spectrogram will indeed be translated onto a Mel scale. The dataset contains information from over 1,600 people from all over the world (1185 men as well as 415 women). The suggested DL model takes as input the CQT as well as Mel-scale spectrograms derived from the breathing and coughing tones of patients diagnosed using the coswara-combined dataset. With the better classification performance employing cough sound CQT and a Mel-spectrogram image, the current proposal outperformed the other nine CNN networks. For patients diagnosed, the accuracy, sensitivity, as well as specificity were 98.9%, 97.3%, and 98.1%, respectively. The Resnet18 is the most reliable network for symptomatic patients using cough and breath sounds. When applied to the Coswara dataset, we discovered that the suggested model's accuracy (98.7%) outperforms the state-of-the-art models (85.6%, 72.9%, 87.1%, and 91.4%) according to the SGDM optimizer. Finally, the research is compared to a comparable investigation. The suggested model is more stable and reliable than any present model. Cough and breathing research precision are good enough just to test extrapolation as well as generalization abilities. As a result, sufferers at their headquarters may utilise this novel method as a main screening tool to try and identify COVID-19 by prioritising patients' RT-PCR testing and decreasing the chance of disease transmission.

Author 1: Mohammed Aly

Author 2: Nouf Saeed Alotaibi

Keywords: COVID-19; median filter; deep learning; Mel-scale spectrogram; sound classification; constant-Q Transform

Paper 70: MFCC and Texture Descriptors based Stuttering Dysfluencies Classification using Extreme Learning Machine

Abstract: Stuttering is a type of speech disorder which results in disrupted flow of speech in the form of unintentional repetitions and prolongation of sounds. Stuttering classification is important for speech pathology treatment and speech therapy techniques which decreases speech disfluency to some extent. In this article, a method for prolongation and repetition classification is presented based on Mel-frequency cepstral coefficients (MFCC) and texture descriptors. Initially, MFCC and filter bank energy (FBE) matrix are computed. Gray level co-occurrence matrix (GLCM) and Gray level run length matrix (GLRLM) textural features are extracted from these matrices. Laplacian score-based feature selection approach is employed to choose relevant features. Finally, extreme learning machine (ELM) is utilized to classify the speech audio event as repetition or prolongation. The algorithm is evaluated using UCLASS database and has achieved improved performance with classification accuracy of 96.36%.

Author 1: Roohum Jegan

Author 2: R. Jayagowri

Keywords: Voice disorder; Mel-frequency cepstral coefficients; gray level co-occurrence matrix; GLRLM; Laplacian score; extreme learning machine

Paper 71: Innovation Management Model as a Source of Business Competitiveness for Industrial SMEs

Abstract: One of the main problems of small companies is not knowing how to properly manage investments, resources, strategies, tools and responsibilities to be more competitive, the research objective is the development of management and innovation processes required by the new company for its permanence in the market and decision making to be carried out every day; small companies face the competitiveness of new products that emerge. Today there are a great variety of products, which are globally interconnected, therefore it is required to implement a structural equation model for its management, so that the company continuously improves and optimizes the available resources, therefore as a result is the management, focused on greater effectiveness of resources; so it is necessary that small businesses need to manage a model of investments, resources, tools and responsibilities to obtain support from the competitiveness of the market; which would allow it to be oriented to a sustainable development and be one step ahead of the competition.

Author 1: Rafael Roosell Paez Advincula

Author 2: Celso Gonzales Chavesta

Author 3: Lilian Ocares-Cunyarachi

Keywords: Competitiveness; continuous improvement; management; optimization; strategies

Paper 72: A Cross Platform Contact Tracing Mobile Application for COVID-19 Infections using Deep Learning

Abstract: The COVID-19 pandemic has remained a global health crisis following the declaration by the World Health Organization. As a result, a number of mechanisms to contain the pandemic have been devised. Popular among these are contact tracing to identify contacts and carry out tests on them in order to minimize the spread of the coronavirus. However, manual contact tracing is tedious and time consuming. Therefore, contact tracing based on mobile applications have been proposed in literature. In this paper, a cross platform contact tracing mobile application that uses deep neural networks to determine contacts in proximity is presented. The application uses Bluetooth Low Energy technologies to detect closeness to a Covid-19 positive case. The deep learning model has been evaluated against analytic models and machine learning models. The proposed deep learning model performed better than analytic and traditional machine learning models during testing.

Author 1: Josephat Kalezhi

Author 2: Mathews Chibuluma

Author 3: Christopher Chembe

Author 4: Victoria Chama

Author 5: Francis Lungo

Author 6: Douglas Kunda

Keywords: Contact tracing mobile application; coronavirus; COVID-19; deep neural networks

Paper 73: A Hybrid 1D-CNN-Bi-LSTM based Model with Spatial Dropout for Multiple Fault Diagnosis of Roller Bearing

Abstract: Fault diagnosis of roller bearings is a crucial and challenging task to ensure the smooth functioning of modern industrial machinery under varying load conditions. Traditional fault diagnosis methods involve preprocessing of the vibration signals and manual feature extraction. This requires domain expertise and experience in extracting relevant features to accurately detect the fault. Hence, it is of great significance to implement an intelligent fault diagnosis method that involves appropriate automatic feature learning and fault identification. Recent research has shown that deep learning is an effective technique for fault diagnosis. In this paper, a hybrid model based on 1D-CNN (One-Dimensional Convolution Neural Networks) with Bi-LSTM (Bi-directional Long-Short Term Memory) is proposed to classify 12 different fault types. Firstly, vibration signals are given as input to 1D-CNN to extract intrinsic features from the input signals. Then, the extracted features are fed into a Bi-LSTM model to identify the faults. The performance of the proposed method is enhanced by applying Softsign activation function in the Bi-LSTM layer and Spatial Dropout in the neural network. To analyze the effectiveness of the proposed method, Case Western Reserve University (CWRU) bearing data is considered for experimentation. The results demonstrated that the proposed model has attained an accuracy of 99.84% in classifying the various faults. The superiority of the proposed method is verified by comparing the predictive accuracy of the proposed method with the existing fault diagnosis methods.

Author 1: Gangavva Choudakkanavar

Author 2: J. Alamelu Mangai

Keywords: Fault diagnosis; roller bearing; deep learning; 1D-CNN; Bi-LSTM; spatial dropout

Paper 74: Building and Testing Fine-Grained Dataset of COVID-19 Tweets for Worry Prediction

Abstract: The COVID-19 outbreak has resulted in the loss of human life worldwide and has increased worry concerning life, public health, the economy, and the future. With lockdown and social distancing measures in place, people turned to social media such as Twitter to share their feelings and concerns about the pandemic. Several studies have focused on analyzing Twitter users’ sentiments and emotions. However, little work has focused on worry detection at a fine-grained level due to the lack of adequate datasets. Worry emotion is associated with notions such as anxiety, fear, and nervousness. In this study, we built a dataset for worry emotion classification called “WorryCov” . It is a relatively large dataset derived from Twitter concerning worry about COVID-19. The data were annotated into three levels (“no-worry”, “worry”, and “high-worry”). Using the annotated dataset, we investigated the performance of different machine learning algorithms (ML), including multinomial Naïve Bayes (MNB), support vector machine (SVM), logistic regression (LR), and random forests (RF). The results show that LR was the optimal approach, with an accuracy of 75%. Furthermore, the results indicate that the proposed model could be used by psychologists and researchers to predict Twitter users’ worry levels during COVID-19 or similar crises.

Author 1: Tahani Soud Alharbi

Author 2: Fethi Fkih

Keywords: COVID-19; sentiment analysis; emotion analysis; worry dataset; concern analysis

Paper 75: Local Pre-Conditioning and Quality Enhancement to Handle Different Data Complexities in Contactless Fingerprint Classification

Abstract: Biometric authentication systems have always been a fascinating approach to meet personalized security. Among the major existing solutions fingerprint-biometrics have gained widespread attention; yet, guaranteeing scalability and reliability over real-time demands remains a challenge. Despite innovations, the recent COVID-19 pandemic has capped the efficacy of the existing touch-based two-dimensional fingerprint detection models. Though, touchless fingerprint detection is considered as a viable alternative; yet, the real-time data complexities like non-linear textural patterns, dusts, non-uniform local conditions like illumination, contrast, orientation make it complex for realization. Moreover, the likelihood of ridge discontinuity and spatio-temporal texture damages can limit its efficacy. Considering these complexities, here, we focused on improving the input image intrinsic feature characteristics. More specifically, applied normalization, ridge orientation estimation, ridge frequency estimation, ridge masking and Gabor filtering over the input touchless fingerprint images. The proposed model mainly focusses on reducing FPR & EER by dividing the input image in to blocks and classify each input block as recoverable and nonrecoverable image block. Finally, an image with higher recoverable blocks with sufficiently large intrinsic features were considered for feature extraction and classification. The Proposed method outperforms when compared with the existing state of the art methods by achieving an accuracy of 94.72%, precision of 98.84%, recall of 97.716%, F-Measure 0.9827, specificity of 95.38% and a reduced EER of about 0.084.

Author 1: Deepika K C

Author 2: G Shivakumar

Keywords: Ridge orientation; Gabor filtering; region masking; ridge frequency; contactless fingerprint

Paper 76: English and Arabic Chatbots: A Systematic Literature Review

Abstract: In recent years, the availability of chatbot applications has increased substantially with the advancement of artificial intelligence techniques, and research efforts have been active in the English language, which presents state-of-the-art solutions. However, despite the popularity of the Arabic language, its research community is still in an immature stage. Therefore, the main objective of this systematic literature review is studying state-of-the-art research – for both the English and Arabic languages – to answer the proposed research questions regarding the development approaches, application domains, evaluation metrics, and development challenges of chatbot applications. The findings show that researchers have devoted more attention to the education domain using retrieval-based approaches while the generation-based approach has grown in popularity recently for providing new responses tasks. Whereas the hybrid approach for ranking multi-possible responses of combining both previous approaches shows a performance improvement. Besides, most metrics used to evaluate chatbot performance are human-based, followed by bilingual evaluation understudy and accuracy metrics. However, defining a common framework for evaluating chatbots remains a challenge. Finally, the open problems and future directions are highlighted to help in developing chatbots with minimal human interference to simulate natural conversations.

Author 1: Abeer S. Alsheddi

Author 2: Lubna S. Alhenaki

Keywords: Chatbots; Arabic language; development approaches; domain applications; evaluation metrics

Paper 77: Duality at Classical Electrodynamics and its Interpretation through Machine Learning Algorithms

Abstract: The aim of this paper is to investigate the hypothetical duality of classical electrodynamics and quantum mechanics through the usage of Machine Learning principles. Thus, the Mitchell’s criteria are used. Essentially this paper is focused on the radiated energy by a free electron inside an intense laser. The usage of mathematical strategies might be correct to some extent so that one expects that classical equation would contain a dual meaning. The concrete case of Compton scattering is analyzed. While at some quantum field theories might not be scrutinized by computer algorithms, contrary to this Quantum Electrodynamics would constitute a robust example.

Author 1: Huber Nieto-Chaupis

Keywords: Classical electrodynamics; quantum mechanics; machine learning principles; mitchell’s criteria

Paper 78: Approximate TSV-based 3D Stacked Integrated Circuits by Inexact Interconnects

Abstract: Three-Dimensional Stacked Integrated Circuit (3D-SICs) based on Through-Silicon Vias (TSVs) provide a high-density integration technology. However, integrating pre-tested dies requires post-bond interconnect testing, which is complex and costly. An imperfect TSV-based interconnect indicates a defective chip that should be rejected. Thus, it increases the yield loss and test cost. On the other hand, approximate computing (AC) is a promising design paradigm suitable for error-resilient applications, e.g., processing sensory-generated data, by judiciously sacrificing output accuracy. AC perform inexact operations and accepts inexact data. Thus, introducing AC into 3D-SICs will significantly ameliorate the efficiency of design approximation. Therefore, this work aims to increase the yield and reduce the test cost by accepting 3D-SICs with defected interconnects as approximate 3D-SICs. This work considers 3D-SICs, where the sensor is stacked on logic (CPU) which is stacked on memory (DRAM). Then, use the memory-based interconnect testing (MBIT) approach to detect and diagnose the faulty interconnect. Based on the detected fault location and type, and for a maximum allowed error, some sensory 3D-SICs with defected LSBs interconnects are accepted and used in error-resilient and data-intensive applications. Targeting data lines only, 50% of the defected interconnects, i.e., least significant bits (LSBs), were accepted as approximate. Thus, the proposed work was able to significantly increase the yield. Two applications, i.e., ECG signal compression and detecting of their R peaks,demonstrated the effectiveness of using a sensory device with a faulty data line in its least significant 8-bits. The approximate ECG signals have a compression rate higher than the exact with negligible (around 0.1%) reduced accuracy.

Author 1: Mahmoud S. Masadeh

Keywords: Approximate computing; Three-Dimensional Inte-grated Circuit (3D IC); Through-Silicon Via (TSV); testing; approx-imate communications; approximate interconnect; yield; energy efficiency

Paper 79: IoT based Low-cost Posture and Bluetooth Controlled Robot for Disabled and Virus Affected People

Abstract: IoT-based robots can help people to a great extent. This work results in a low-cost posture recognizer robot that can detect posture signs from a disabled or virus-affected person and move accordingly. The robot can take images with the Raspberry Pi camera and process the image to identify the posture with our designed algorithm. In addition, it can also take instructions via Bluetooth from smartphone apps. The robot can move 360 degrees depending on the input posture or Bluetooth. This system can assist disabled people who can move a few organs only. Moreover, this system can assist virus-affected persons as they can instruct the robot without touching it. Finally, the robot can collect data from a distant place and send it to a cloud server without spreading the virus.

Author 1: Tajim Md. Niamat Ullah Akhund

Author 2: Mosharof Hossain

Author 3: Khadizatul Kubra

Author 4: Nurjahan

Author 5: Alistair Barros

Author 6: Md Whaiduzzaman

Keywords: Internet of Things (IoT); Raspberry Pi; Pi camera; Pi robots; hospital robots; posture recognition robots

Paper 80: Deepfakes on Retinal Images using GAN

Abstract: In Deep Learning (DL), Generative Adversarial Networks (GAN) are a popular technique for generating synthetic images, which require extensive and balanced datasets to train. These Artificial Intelligence systems can produce synthetic images that seem authentic, known as Deep Fakes. At present, data-driven approaches to classifying medical images are prevalent. However, most medical data is inaccessible to general researchers due to standard consent forms that restrict research to medical journals or education. Our study focuses on GANs, which can create artificial fundus images that can be indistinguishable from actual fundus images. Before using these fake images, it is essential to investigate privacy concerns and hallucinations thoroughly. As well as, reviewing the current applications and limitations of GANs is very important. In this work, we present the Cycle-GAN framework, a new GAN network for medical imaging that focuses on the generation and segmentation of retinal fundus images.DRIVE retinal fundus image dataset is used to evaluate the proposed model’s performance and achieved an accuracy of 98.19%.

Author 1: Yalamanchili Salini

Author 2: J HariKiran

Keywords: DeepFakes; deep learning; retinal fundus image synthesis; segmentation; generative adversarial network (GAN); variational autoencoder (VAE)

Paper 81: A GIS and Fuzzy-based Model for Identification and Analysis of Accident Vulnerable Locations in Urban Traffic Management System: A Case Study of Bhubaneswar

Abstract: The world has seen road accident and its related societal and economical impact as one of the live, persisting problem in the last 2-3 decades and its prominence has been observed in the developing countries of the Asian subcontinent. With no exception, all major cities in India are facing the various challenges related to road accidents, mostly due to the large population density. Among the major cities in India, Bhubaneswar is a very fast growing city with aim to be the most livable city in Asia in the coming few years. With the city of Bhubaneswar as our study area, we address the issues related to road safety by determining the degree of severity of road accidents. We study the accident related data collected for the last decade using the spatial tools of Geographical Information System (GIS). Then using a GIS-map based analysis and a fuzzy-based model, we have found the spatio-temporal distribution of accident vulnerable locations with their degree of severity. Our experimental results show the accident hot-spots with values of selected contributing parameters such as timing, traffic density, vehicle speed, road intersections.

Author 1: Sarita Mahapatra

Author 2: Krishna Chandra Rath

Author 3: Satya Ranjan Das

Keywords: Road traffic management; accident vulnerability; GIS; fuzzy inference; fuzzy rules

Paper 82: Novel Deep Learning Technique to Improve Resolution of Low-Quality Finger Print Image for Bigdata Applications

Abstract: High-resolution images are highly in demand when they are utilized for different analysis purposes and obviously due to their quality aesthetic visual impact. The objective of image super-resolution (SR) is to reconstruct a high-resolution (HR) image from a low-resolution (LR)image. Storing, transferring and processing of high-resolution (HR) images have got many practical issues in big data domain. In the case of finger print images, the data is huge because of the huge number of populations. So instead of transferring or storing these finger print images in its original form (HR images), it cost very low if we choose its low-resolution form. By using sampling technique, we can easily generate LR images, but the main problem is to regenerate HR image from these LR images. So, this paper addresses this problem, a novel method for enhancing resolution of low-resolution fingerprint images of size 50 x 50 to a high-resolution image of size 400 x 400 using convolutional neural network (CNN) architecture followed by sub pixel convolution operation for up sampling with no loss of promising features available in low-resolution image has been proposed. The pro-posed model contains five convolutional layers, each of which has an appropriate number of filter channels, activation functions, and optimization functions. The proposed model was trained using three publicly accessible fingerprint datasets FVC 2004 DB1, DB2, and DB3 after being validation and testing were done using 10 percent of these fingerprint data sets. In terms of performance measures like Peak Signal to Noise Ratio (PSNR), Mean Squared Error (MSE), Structural Similarity Index (SSIM) and loss functions, the quantitative and qualitative results show that the proposed model greatly outperformed existing state-of-the-art techniques like Enhanced deep residual network (EDSR), wide activation for image and video SR (WDSR), Generative adversarial network(GAN) based models and Auto-encoder-based models.

Author 1: Lisha P P

Author 2: Jayasree V K

Keywords: Single image super-resolution; convolution neural network; biometric; fingerprint images

Paper 83: Mobile Application: A Proposal for the Inventory Management of Pharmaceutical Industry Companies

Abstract: In recent years, the development of mobile applications has been evolving and becoming more and more frequent. This event is positive, since it plays an important role in facing and mitigating the multiple adversities that appear in the different existing sectors, such as business. On the other hand, it was detected that a little known problem that many companies in the pharmaceutical industry experience is poor inventory management, which causes countless consequences, generally of a negative nature. For this reason, in this work it was decided to make a mobile application prototype to face this problem. In this regard, the RUP methodology was used, along with various computer tools, in order to elaborate the prototype. Besides, as a data collection technique, surveys were made, which were subjected to expert judgment, in order to qualify the prototype. Likewise, very satisfactory results were obtained, concluding that the mobile application prototype that was developed complies with all the necessary conditions to mitigate the inventory management problems of pharmaceutical industry companies.

Author 1: Alfredo Leonidas Vasquez Ubaldo

Author 2: Juan Andres Berrios Albines

Author 3: Jose Luis Herrera Salazar

Author 4: Laberiano Andrade-Arenas

Author 5: Michael Cabanillas-Carbonell

Keywords: Inventory management; mobile application; pharmaceutical industry; prototype; RUP methodology

Paper 84: A Deep Neural Network based Detection System for the Visual Diagnosis of the Blackberry

Abstract: Thanks to its geographical and climatic advantages, Colombia has a historically strong fruit-growing tradition. To date, the basis of its food and economic development in a significant part of its territory is based on a wide range of fruits. One of the most important in the central and western regions of the country is the blackberry, which is rooted not only from the economic and food point of view but also culturally. For the departments of Casanare, Santander, and Cundinamarca, this fruit is one of the primary sources of income, rural employment, and food supply and income. However, small and medium farmers cultivate without access to technological production tools and with limited economic capacity. This process suffers from several problems that affect the whole plant, especially the fruit, which is strongly influenced by fungi, extreme ripening processes, or low temperatures. One of the main problems to be dealt with in its cultivation is the spread of pests, which are one of the causes of fruit rot. As a support strategy in producing this fruit, the development of an embedded system for visually diagnosing the fruit using a deep neural network is proposed. The article presents the training, tuning, and performance evaluation of this convolutional network to detect three possible fruit states, ripe, immature, and rotten, to facilitate the harvesting and marketing processes and reduce the impact on the healthy fruit and the quality of the final product. The model is built with a ResNet type network, which is trained with its dataset, which seeks to use images captured in their natural environment with as little manipulation as possible to reduce image analysis. This model achieves an accuracy of 70%, which indicates its high performance and validates its use in a stand-alone embedded system.

Author 1: Alejandro Rubio

Author 2: Carlos Avendano

Author 3: Fredy Martinez

Keywords: Automatic sorter; blackberry; deep neural network; fruit handling; image analysis

Paper 85: A Comparative Analysis of Generative Neural Attention-based Service Chatbot

Abstract: Companies constantly rely on customer support to deliver pre-and post-sale services to their clients through websites, mobile devices or social media platforms such as Twitter. In assisting customers, companies employ virtual service agents (chatbots) to provide support via communication devices. The primary focus is to automate the generation of conversational chat between a computer and a human by constructing virtual service agents that can predict appropriate and automatic responses to customers’ queries. This paper aims to present and implement a seq2seq-based learning task model based on encoder-decoder architectural solutions by training generative chatbots on customer support Twitter datasets. The model is based on deep Recurrent Neural Networks (RNNs) structures which are uni-directional and bi-directional encoder types of Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU). The RNNs are augmented with an attention layer to focus on important information between input and output sequences. Word level embedding such as Word2Vec, GloVe, and FastText are employed as input to the model. Incorporating the base architecture, a comparative analysis is applied where baseline models are compared with and without the use of attention as well as different types of input embedding for each experiment. Bilingual Evaluation Understudy (BLEU) was employed to evaluate the model’s performance. Results revealed that while biLSTM performs better with Glove, biGRU operates better with FastText. Thus, the finding significantly indicated that the attention-based, bi-directional RNNs (LSTM or GRU) model significantly outperformed baseline approaches in their BLEU score as a promising use in future works.

Author 1: Sinarwati Mohamad Suhaili

Author 2: Naomie Salim

Author 3: Mohamad Nazim Jambli

Keywords: Sequence-to-sequence; encoder-decoder; service chatbot; attention-based encoder-decoder; Recurrent Neural Net-work (RNN); Long Short-Term Memory (LSTM); Gated Recurrent Unit (GRU); word embedding

Paper 86: CapNet: An Encoder-Decoder based Neural Network Model for Automatic Bangla Image Caption Generation

Abstract: Automatic caption generation from images has be-come an active research topic in the field of Computer Vision (CV) and Natural Language Processing (NLP). Machine generated image caption plays a vital role for the visually impaired people by converting the caption to speech to have a better understanding of their surrounding. Though significant amount of research has been conducted for automatic caption generation in other languages, far too little effort has been devoted to Bangla image caption generation. In this paper, we propose an encoder-decoder based model which takes an image as input and generates the corresponding Bangla caption as output. The encoder network consists of a pretrained image feature extractor called ResNet-50, while the decoder network consists of Bidirectional LSTMs for caption generation. The model has been trained and evaluated using a Bangla image captioning dataset named BanglaLekhaIm-ageCaptions. The proposed model achieved a training accuracy of 91% and BLEU-1, BLEU-2, BLEU-3, BLEU-4 scores of 0.81, 0.67, 0.57, and 0.51 respectively. Moreover, a comparative study for different pretrained feature extractors such as VGG-16 and Xception is presented. Finally, the proposed model has been deployed on an embedded device for analysing the inference time and power consumption.

Author 1: Rashik Rahman

Author 2: Hasan Murad

Author 3: Nakiba Nuren Rahman

Author 4: Aloke Kumar Saha

Author 5: Shah Murtaza Rashid Al Masud

Author 6: A S Zaforullah Momtaz

Keywords: Bangla image caption generation; encoder-decoder; bidirectional long short term memory (LSTM); bangla natural language processing (NLP)

Paper 87: An Improved K-Nearest Neighbor Algorithm for Pattern Classification

Abstract: This paper proposed a “Locally Adaptive K-Nearest Neighbor (LAKNN) algorithm” for pattern exploration problem to enhance the obscenity of dimensionality. To compute neighborhood local linear discriminant analysis is an effective metric which determines the local decision boundaries from centroid information. KNN is a novel approach which uses in many classifications problem of data mining and machine learning. KNN uses class conditional probabilities for unfamiliar pattern. For limited training data in high dimensional feature space this hypothesis is unacceptable due to disfigurement of high dimensionality. To normalize the feature value of dissimilar metrics, Standard Euclidean Distance is used in KNN which s misguide to find a proper subset of nearest points of the pattern to be predicted. To overcome the effect of high dimensionality LANN uses a new variant of Standard Euclidian Distance Metric. A flexible metric is estimated for computing neighborhoods based on Chi-squared distance analysis. Chi-squared metric is used to ascertains most significant features in finding k-closet points of the training patterns. This paper also shows that LANN outperformed other four different models of KNN and other machine-learning algorithm in both training and accuracy.

Author 1: Zinnia Sultana

Author 2: Ashifatul Ferdousi

Author 3: Farzana Tasnim

Author 4: Lutfun Nahar

Keywords: LANN algorithm; Standard Euclidian Distance; variance based Euclidian Distance; feature extraction; pattern classification

Paper 88: An Approach to Detect Phishing Websites with Features Selection Method and Ensemble Learning

Abstract: Nowadays, phishing is a major problem on a global scale. Everyone must use the internet in today’s society in order to cope up in the real world. As a result, internet crime like phishing has become a serious issue throughout the world. This type of crime can be committed by anyone; all they need is a computer. Additionally, hacking may now be learned quickly by anyone with programming and mathematical skills. The adoption of various techniques by anti-phishing toolbars, such as machine learning, may enable users to quickly identify a fake website. As a result, researchers are now particularly interested in the problem of detecting fraudulent websites. Machine learning techniques have been offered throughout the entire process to more precisely identify fraudulent websites. To find the best accurate outcome, classification with random parameter tuning and ensemble based approaches are utilized. A user-friendly interface has also been suggested to make the system more accessible to the public.

Author 1: Mahmuda Khatun

Author 2: MD Akib Ikbal Mozumder

Author 3: Md. Nazmul Hasan Polash

Author 4: Md. Rakib Hasan

Author 5: Khalil Ahammad

Author 6: MD. Shibly Shaiham

Keywords: Machine learning; deep learning; catboost; LGBM; embedded; react-native; flask

Paper 89: A Prototype Implementation of a CUDA-based Customized Rasterizer

Abstract: In these days, we have high-performance massively parallel computing devices, as well as high-performance 3D graphics rendering devices. In this paper, we show a prototype implementation of a full-software 3D rasterizer system, based on the CUDA parallel architecture. While most of previous CUDA-based software rasterizer implementations focused on the triangle primitives, our system includes more 3D primitives, and extra 2D primitives, to fully support 3D graphics library features. Currently, our system is at its prototype implementation stage, and it shows successful results with 3D primitive handling and also character output features. Our design and implementation details are presented. More optimizations and fine tunes will be followed in near future.

Author 1: Nakhoon Baek

Keywords: 3D rasterization; CUDA implementation; OpenGL emulation

Paper 90: Mobile App Design: Logging and Diagnostics of Respiratory Diseases

Abstract: Over the years, a wide variety of respiratory diseases have caused a high mortality rate throughout the world. This was again observed with the appearance of the pandemic, COVID-19. In addition, the most affected are people living in extreme poverty. The objective is design a mobile health application for the registration and diagnosis of respiratory diseases. For this, the RUP methodology was applied, because it easily adapts to various types of projects. Its use, together with the UML process development software, allows the analysis, implementation and documentation of object oriented systems. For validation, a user survey was carried out and the questionnaire was based on the dimensions of functionality, efficiency, effectiveness and satisfaction. Obtaining as a result a positive qualification to the design of the application and its acceptance due to the reduction in the time to obtain the diagnosis. In conclusion, a mobile health application design was successfully carried out so that patients can register and have the diagnosis of respiratory diseases from the comfort of their home.

Author 1: Diana Cecilia Chavez Canari

Author 2: Angel Vicente Garcia Obispo

Author 3: Jose Luis Herrera Salazar

Author 4: Laberiano Andrade-Arenas

Author 5: Michael Cabanillas-Carbonell

Keywords: Mobile app; Covid-19; diagnosis; respiratory diseases; RUP methodology

Paper 91: Summarizing Event Sequence Database into Compact Big Sequence

Abstract: Detecting the core structure of a database is one of the most objective of data mining. Many methods do so, in pattern set mining, by mining a small set of patterns that together summarize the dataset in efficient way. The better of these patterns, the more effective summarization of the database. Most of these methods are based on the Minimum Description Length principle. Here, we focus on the event sequence database. In this paper, rather than mining a small set of significant patterns, we propose a novel method to summarize the event sequence dataset by constructing compact big sequence namely, BigSeq. BigSeq conserves all characteristics of the original event sequences. It is constructed in efficient way via the longest common subsequence and the novel definition of the compatible event set. The experimental results show that BigSeq method outperforms the state-of-the-art methods such as Gokrimp with respect to compression ratio, total response time, and number of detected patterns.

Author 1: Mosab Hassaan

Keywords: Sequence data; compressing patterns mining; minimum description length

Paper 92: A Novel Big Data Intelligence Analytics Framework for 5G-Enabled IoT Healthcare

Abstract: Intelligent networking is a concept that enables 5G, the Internet of Things (IoT) and artificial intelligence (AI) to combine as a way to accelerate technological innovation and develop new revolutionary digital services. In the intelligent connectivity vision, the digital information gathered by the machines, devices and sensors which make up the IoT is analysed and contextualized. It is anticipated that the high availability of 5G and its inclusion of a large number of connections would help promote the production of wearable devices used to monitor the different biometric parameters of the wearer. Since these are AIbased health systems, the data obtained from these devices will be analysed in order to assess a patient’s current health status. This paper presents a detailed design for the development of intelligent data analytics and mobile computer-assisted healthcare systems.The proposed advanced PoS consensus algorithm provides better performance than other existing algorithms.

Author 1: Yassine Sabri

Author 2: Ahmed Outzourhit

Keywords: Big data analytics; 5G-enabled; IoT healthcare; fog computing; confidentiality

Paper 93: A New Learning to Rank Approach for Software Defect Prediction

Abstract: Software defect prediction is one of the most active research fields in software development. The outcome of defect prediction models provides a list of the most likely defect-prone modules that need a huge effort from quality assurance teams. It can also help project managers to effectively allocate limited resources to validating software products and invest more effort in defect-prone modules. As the size of software projects grows, error prediction models can play an important role in assisting developers and shortening the time it takes to create more reliable software products by ranking software modules based on their defects. Therefore, there is need a learning-to-rank approach that can prioritize and rank defective modules to reduce testing effort, cost, and time. In this paper, a new learning to rank approach was developed to help the QA team rank the most defect-prone modules using different regression models. The proposed approach was evaluated on a set of standardized datasets using well-known evaluation measures such as Fault-Percentile Average (FPA), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and the Cumulative Lift Chart (CLC). Also, our proposed approach was compared with some other regression models that are used for software defect prediction, such as Random Forest (RF), Logistic Regression (LR), Support Vector Regression (SVR), Zero Inflated Regression (ZIR), Zero Inflated Poisson (ZIP), and Negative Polynomial Regression (NPR). Based on the results, the measurement criteria were different than each other as there was a gap in the accuracy obtained for defects prediction due to the nature of the random data, and thus was higher for RF and SVR, as well as FPA achieved better results than MAE and RMSE in this research paper.

Author 1: Sara Al-omari

Author 2: Yousef Elsheikh

Author 3: Mohammed Azzeh

Keywords: Software engineering; software testing; software defect prediction; learning to rank approach

Paper 94: Utilizing Artificial Intelligence Techniques for Assisting Visually Impaired People: A Personal AI-based Assistive Application

Abstract: Nowadays, the Artificial Intelligence (AI) field has made a significant change in the real life. Numerous applications use the AI techniques for the purpose of assisting people in different life aspects. Furthermore, with the increased number of people with visual difficulties around the world, there is a need for such AI assistive applications which provide them an independent life. Limited affordable and appropriate solutions developed so far. In this paper, we present a personal AI-based assistive application called (Vivid) that supports visually impaired people being more independent. Vivid has many features such as identifying objects, objects’ colors, recognizing text, and faces detection. It relies on using the mobile camera to sense the environment, and the machine learning techniques to understand the environment. By translating a meaningful information in audible sound for those users, Vivid does not require to have any visual ability. Moreover, the whole interaction with the user is only based on voice commands. The input from the user is captured as finger gestures on tablet or cell phone touch screen. In addition to Vivid, we also shade the lights on a supplementary application that notify/alarm visually impaired people of any nearby objects using sensors. These personal assistive applications were developed then tested on the real world and showed promising results.

Author 1: Samah Alhazmi

Author 2: Mohammed Kutbi

Author 3: Soha Alhelaly

Author 4: Ulfat Dawood

Author 5: Reem Felemban

Author 6: Entisar Alaslani

Keywords: Artificial intelligence; machine learning; assistive technology; visually impaired

Paper 95: Fashion Image Retrieval based on Parallel Branched Attention Network

Abstract: With the increase in vision-associated applications in e-commerce, image retrieval has become an emerging application in computer vision. Matching the exact user clothes from the database images is challenging due to noisy background, wide variation in orientation and lighting conditions, shape deformations, and the variation in the quality of the images between query and refined shop images. Most existing solutions tend to miss out on either incorporating low-level features or doing it effectively within their networks. Addressing the issue, we propose an attention-based multiscale deep Convolutional Neural Network (CNN) architecture called Parallel Attention ResNet (PAResNet50). It includes other supplementary branches with attention layers to extract low-level discriminative features and uses both high-level and low-level features for the notion of visual similarity. The attention layer focuses on the local discriminative regions and ignores the noisy background. Image retrieval output shows that our approach is robust to different lighting conditions. Experimental results on two public datasets show that our approach effectively locates the important region and significantly improves retrieval accuracy over simple network architectures without attention.

Author 1: Sangam Man Buddhacharya

Author 2: Sagar Adhikari

Author 3: Ram Krishna Lamichhane

Keywords: Convolutional neural network (CNN); image retrieval; attention mechanism; convolutional block attention module (CBAM)

Paper 96: Watchdog Monitoring for Detecting and Handling of Control Flow Hijack on RISC-V-based Binaries

Abstract: Control flow hijacking has been a major challenge in software security. Several means of protections have been developed but insecurities persist. This is because existing protections have sometimes been circumvented while some resilient protections do not cover all applications. Studies have revealed that a holistic way of tackling software insecurity could involve watchdog monitoring and detection via Control Flow Integrity (CFI). The CFI concept has shown a good measure of reliability to mitigate control flow hijacking. However, sophisticated attack techniques in the form of Return Oriented Programming (ROP) have persisted. A flexible protection is desirable, which not only covers as many architecture structures as possible but also mitigates known resilient attacks like ROP. The solution proffered here is a hybrid of CFI and watchdog timing via inter-process signaling (IP-CFI). It is a software-based protection that involves recompilation of the target program. The implementation here is on vulnerable RISC-V-based process but is flexible and could be adapted on other architectures. We present a proof of concept in IP-CFI which when applied to a vulnerable program, ROP is mitigated. The target program incurs a run-time overhead of 1.5%. The code is available.

Author 1: Toyosi Oyinloye

Author 2: Lee Speakman

Author 3: Thaddeus Eze

Author 4: Lucas O’Mahony

Keywords: Watchdog; return oriented programming; RISC-V; control flow integrity; software security

Paper 97: Towards Flexible Transparent Authentication System for Mobile Application Security

Abstract: Undoubtedly, Mobile Application Security (MAS) has made tremendous progress in implementing enhanced security protocols in the past decade. With the recent increase in the usage of mobile applications, concerns of privacy and security are increasing rapidly. Thus, the security measurement must be applied to satisfy security and privacy needs. On the one hand, the developer community works feverishly to develop mobile applications with innovative and usable layers and user-friendly for multigenerational customers. However, the security community, in particular, strives to make those layers secure. Therefore, the main objective of this research is to build a transparent authentication system in a mobile application. There are potentially many ways to implement an authentication mechanism such as the biometrics approach. It has features, which can be used to heightened security for the end-user. In these articles, we experimentally investigate the multigenerational customer base’s factors such as age, convenience, easiness, memorizing new passwords, and understanding the precept of frequently changing passwords to enhance security. Additionally, we propose a system that will solve the common problems users face when starting the password resetting process. At the same time, in the MAS sector, we orchestrate the applications for better security encryption for the stored biometrics to ensure it, which makes it even more challenging for an adversary to bypass the system and reset the password. We conclude our research with a comprehensive security solution for MAS that considers user friendliness and data safeguarding.

Author 1: Abdullah Golam

Author 2: Mohammed Abuhmoud

Author 3: Umar Albalawi

Keywords: Transparent security; authentication; UX/UI; for-getting password; reset password; biometric systems

Paper 98: Computational Analysis based on Advanced Correlation Automatic Detection Technology in BDD-FFS System

Abstract: Big Data-Driven Fabric Future Systems (BDD-FFS) is currently attracting widespread attention in the healthcare research community. Medical devices rely primarily on the intelligent Internet to gather important health-related information. According to this, we provide patients with deeply supportive data to help them through their recovery. However, due to the large number of medical devices, the address of the device can be modified by intruders, which can be life-threatening for serious patients (such as tumor patients). A large number of abnormal cells in the brain can lead to brain tumors, which harm brain tissue and can be life-threatening. Recognition of brain tumors at the beginning of the process is significant for their detection, prediction and therapy. The traditional approach for detecting is for a human to perform a biopsy and examine CT scans or magnetic resonance imaging (MRI), which is cumbersome, unrealistic for great amounts of resource, and requests the radiologist to make inferential computations. A variety of automation schemes have been designed to address these challenges. However, there is an urgent need to develop a technology that will detect brain tumors with remarkable accuracy in a much shorter time. In addition, the selection of feature sets for prediction is crucial to realize significant accuracy. This work utilizes an associative action learner with an advanced feature group, Partial Tree (PART-T), to detect brain tumor recognition grades. The model presented was compared with existing methods through 10-fold cross-validation. Experimental results show that partial trees with advanced feature sets are superior to existing techniques in terms of performance indicators used for evaluation, such as accuracy, recall rate and F-measure.

Author 1: Xiao Zheng

Author 2: Muhammad Tahir

Author 3: Mingchu Li

Author 4: Shaoqing Wang

Keywords: Big data-driven fabric future systems (BDD-FFS); magnetic resonance imaging (MRI); partial tree

Paper 99: Image Enhancement Method based on an Improved Fuzzy C-Means Clustering

Abstract: Image enhancement is an important method in the process of image processing. This paper proposes an image enhancement method base on an improved fuzzy c-means clustering. The method consists of the following steps: firstly, proposed a fuzzy c-means clustering with a cooperation center(FCM-co). Secondly, using the FCM-co, divide the image pixels into different clusters and marked membership values to those clusters. Thirdly, modify the membership values. Finally, calculate the new pixel gray levels. This enhancement method can overcome the disadvantage of overexposure and better retain image details. Through the experiment, the test results show that the proposed enhancement method could achieve better performance.

Author 1: Libao Yang

Author 2: Suzelawati Zenian

Author 3: Rozaimi Zakaria

Keywords: Image enhancement; fuzzy clustering; fuzzy c-means clustering; membership; objective function

Paper 100: A New Hate Speech Detection System based on Textual and Psychological Features

Abstract: Hate speech often spreads on social media and harms individuals and the community. Machine learning models have been proposed to detect hate speech in social media; however, several issues presently limit the performance of current approaches. One challenge is the issue of having diverse comprehensions of hate speech constructs which will lead to many speech categories and different interpretations. In addition, certain language-specific features, and short text issues, such as Twitter, exacerbate the problem. Moreover, current machine learning approaches lack universality due to small datasets and the adoption of a few features of hateful speech. This paper develops and builds new feature sets based on frequencies of textual tokens and psychological characteristics. Then, the study evaluates several machine learning methods over a large dataset. Results showed that the Random Forest and BERT methods are the most valuable for detecting hate speech content. Furthermore, the most dominant features that are helpful for hate speech detection methods combine psychological features and Term-Frequency Inverse Document-Frequency (TFIDF) features. Therefore, the proposed approach could identify hate speech on social media platforms like Twitter.

Author 1: Fatimah Alkomah

Author 2: Sanaz Salati

Author 3: Xiaogang Ma

Keywords: Hate speech detection; hate speech classification; hate speech features; hate speech methods

Paper 101: Straggler Mitigation in Hadoop MapReduce Framework: A Review

Abstract: Processing huge and complex data to obtain useful information is challenging, even though several big data processing frameworks have been proposed and further enhanced. One of the prominent big data processing frameworks is MapReduce. The main concept of MapReduce framework relies on distributed and parallel processing. However, MapReduce framework is facing serious performance degradations due to the slow execution of certain tasks type called stragglers. Failing to handle stragglers causes delay and affects the overall job execution time. Meanwhile, several straggler reduction techniques have been proposed to improve the MapReduce performance. This study provides a comprehensive and qualitative review of the different existing straggler mitigation solutions. In addition, a taxonomy of the available straggler mitigation solutions is presented. Critical research issues and future research directions are identified and discussed to guide researchers and scholars.

Author 1: Lukuman Saheed Ajibade

Author 2: Kamalrulnizam Abu Bakar

Author 3: Ahmed Aliyu

Author 4: Tasneem Danish

Keywords: Big data; blacklisting execution; Hadoop; MapReduce; spark; speculative execution; straggler

Paper 102: Blood Management System based on Blockchain Approach: A Research Solution in Vietnam

Abstract: More and more new health care solutions are born based on the development of science and technology. The subjects who benefit the most, in this case, are not only patients (i.e., shorter healing time, faster recovery) but also medical staff, e.g., doctors/nurses (i.e., easy monitoring of the patient’s recovery process, proposing new treatment). However, there are still products that have not found an alternative: blood and blood products. Regardless of how science and technology can affect all aspects of patient treatment as well as medical care, blood still plays an important role in the treatment method. In addition to the above, blood and blood products may only be obtained from volunteers (i.e., blood donors). The preservation process is also very difficult, and no medical facility has enough facilities to preserve them. Therefore, the current process of blood preservation and transportation is done manually and contains many potential risks (e.g., data loss, personal information collection). In addition to the above barriers, developing countries (including Vietnam) also face many difficulties due to limited facilities. It is for this reason that this paper aims at a Blockchain-based technology solution for efficient management and distribution of blood from blood products. On the one hand, the paper contributes to the limitations in the information management process of storing and transporting blood and its products in the traditional database being applied in medical facilities in the cities and provinces in the Mekong Delta (the West-South of Vietnam). On the other hand, the article offers technology-based solutions to increase transparency and reduce the fear of centralized data storage (i.e., security and privacy issues). We also implement a proof-of-concept to evaluate the feasibility of the proposed approach.

Author 1: Hieu Le Van

Author 2: Hong Khanh Vo

Author 3: Luong Hoang Huong

Author 4: Phuc Nguyen Trong

Author 5: Khoa Tran Dang

Author 6: Khiem Huynh Gia

Author 7: Loc Van Cao Phu

Author 8: Duy Nguyen Truong Quoc

Author 9: Nguyen Huyen Tran

Author 10: Huynh Trong Nghia

Author 11: Bang Le Khanh

Author 12: Kiet Le Tuan

Keywords: Blood donation in Vietnam; blockchain; hyperledger fabric; blood products supply chain

Paper 103: Letter-of-Credit Chain: Cross-Border Exchange based on Blockchain and Smart Contracts

Abstract: The exchange of goods between countries is growing, contributing to the promotion of logistics-related technologies. More and more systems are adopting advances in science and engineering to reduce manual handling steps, thereby reducing transit time. Letter-of-Credit (LOC) is a standard method where the parties involved will enter into agreements for the sale and exchange of goods. Specifically, each party will receive a set of original documents and does not need to meet face-to-face under the bank’s witness. The process brings many benefits in terms of time and reduces records processing. However, the system faces a lot of risks when one of the parties is dishonest. On the other hand, the traditional LOC systems face a lot of risks related to the transparency of information about the goods, and also the supplier may lose the goods (e.g., 4/100 Vietnamese cashew nut containers are lost. stuck in Italy) or deposits in the hands of shipping companies (e.g., GNN Express - Vietnam) and many more. To this end, many research directions have exploited blockchain technology and smart contracts. Specifically, all information related to the transaction between the supplier and the demander including package, time, and delivery location. However, there needs to be a mechanism to ensure the smooth implementation of smart contracts, specifically for sanctioning when there is a conflict between a supplier and a demander. This role should be considered for the transaction manager, who directly designs and is responsible for their smart contracts. Currently, there is no mechanism to guarantee all interests of the parties involved in non-bank transactions. To increase the processing capacity and integrate with the Blockchain system, we propose the Letter-of-credit Chain that defines the agreements between the parties in international trade. We also deploy the proof-of-concept of the Letter-of-credit Chain on the three EVM-supported platforms (i.e., under ERC20), namely, Ethereum, Binance Smart Chain, and Fantom. By evaluating the actual execution of Gas for each platform, we found that our proposed model had the cheapest fee when deployed on the Fantom platform. Finally, we share the deployment/implementation of these platforms’ proof-of-concept to encourage further future research.

Author 1: Khoi Le Quoc

Author 2: Phuc Nguyen Trong

Author 3: Hieu Le Van

Author 4: Hong Khanh Vo

Author 5: Luong Hoang Huong

Author 6: Khoa Tran Dang

Author 7: Khiem Huynh Gia

Author 8: Loc Van Cao Phu

Author 9: Duy Nguyen Truong Quoc

Author 10: Nguyen Huyen Tran

Author 11: Huynh Trong Nghia

Author 12: Bang Le Khanh

Author 13: Kiet Le Tuan

Keywords: Letter-of-Credit; blockchain; smart contract; authorization; Ethereum; Fantom; Binance smart chain

Paper 104: Enhanced Security: Implementation of Hybrid Image Steganography Technique using Low-Contrast LSB and AES-CBC Cryptography

Abstract: Now-a-days, sensitive and confidential information needs to be exchanged over open, public, and not secure networks such as the Internet. For this purpose, some information security techniques combine cryptographic and steganographic algorithms and image processing techniques to exchange information securely. Therefore, this research presents the implementation of an algorithm that combines the AES-CBC cryptographic technique with the LSB steganographic technique, which is statistically enhanced by image processing by looking for low-contrast areas where the encrypted information will be stored. This hybrid algorithm was developed to send a plaintext file hidden in an image in BMP format, so the changes in the image are invisible to the human eye and undetectable in possible steganographic analysis. The implementation was performed using Python and its libraries PyCryptodome for encryption and CV2 for image processing. As a result, it was found that the hybrid algorithm implemented has three layers of security over a plaintext encrypted and hidden in a digital image, which makes it difficult to break the secrecy of the information exchanged in a stego-image file. Additionally, the execution times of the hybrid algorithm were evaluated for different sizes of plaintext and digital image files.

Author 1: Edwar Jacinto G

Author 2: Holman Montiel A

Author 3: Fredy H. Martínez S

Keywords: Steganography; cryptography; LSB; low contrast areas; AES-CBC algorithm

Paper 105: Secure and Efficient Implicit Certificates: Improving the Performance for Host Identity Protocol in IoT

Abstract: Implicit certificates own the shorter public key validation data. This property makes them appealing in resource-constrained IoT systems where public-key authentication is performed very often, which is common in Host Identity Protocol (HIP). However, it is still a critical challenge in IoT how to guarantee the security and efficiency of implicit certificates. This article presents a forgery attack for the Privacy-aware HIP (P-HIP), and then propose a Secure and Efficient Implicit Certificate (SEIC) scheme that can improve the security of the P-HIP and the efficiency of elliptic-curve point multiplications for IoT devices. For a fix-point multiplication, the proposed approach is about 1.5 times faster than the method in SIMPL scheme. Furthermore, we improve the performance of SEIC with the butterfly key expansion process, and then construct an improved P-HIP. Experimental results show that compare to the existing schemes, the improved scheme makes a user/device have both the smallest computation cost and the smallest communication cost.

Author 1: Zhaokang Lu

Author 2: Jianzhu Lu

Keywords: Authentication; privacy; implicit certificates; internet of things (IoT); host identity; security