Journal of Informatics and Web Engineering

Editorial Preview for June 2025 Issue

2025-06-13T12:34:34+08:00

Effective from Volume 3, JIWE has transitioned to a triannual publication release effort. Specifically, releases occur each February, June, and October. This June 2025 release contains 28 papers within the regular section that covers a broad range of application concerning Machine Learning (ML), Augmented Intelligence, Artificial Intelligence (AI), Generative AI, Data Mining (DM), Software Engineering, Sentiment Analysis, Recommender Systems, Healthcare, and other key areas in web engineering. Additionally, this edition presents a captivating collection of 6 papers curated by our Thematic Editor, Assoc. Prof. Dr. Heru Agus Santoso, under the theme "Augmented Intelligence". In his editorial, Assoc. Prof. Dr. Heru Agus Santoso highlights cutting-edge research on the integration of software engineering and business intelligence to support knowledge-driven strategies for competitive advantage. Moreover, these papers are aligned to some of the United Nations Sustainable Development Goals (SDGs), particularly SDG 3 (Good Health and Well-being) through advancements in healthcare technologies, SDG 9 (Industry, Innovation, and Infrastructure) through software and systems innovation research, and SDG 16 (Peace, Justice, and Strong Institutions) through contributions to privacy and cybersecurity research.

Improving Culinary Informatics through Meaningful Social Web Engineering

2025-01-12T17:43:31+08:00

This paper gives a good reflection about social web engineering aspirations to ardently see the truth to enlighten culinary concentration and wisdom. With the compassionate path to discover food and engage customers to share content with loving kindness spread to every corner. There are so many features realized, which include direct interfaces through intuitive interfaces to lead the daily food streamlining. May one engaged with constructive creation to grow compassions in every culinary living sharing. Contentment would develop in our heart involved in different phases put forward in this paper. The initial phase begins with honesty in planning and design. Next, the right livelihood of background study brings about requirements definition. This cast out diligently the interactive user conceptualization to arouse wholesome of good and wise design diagrams. With this, the users can reach the goal of registering and establishing profiles. This practice leads to steady mindful culinary owners as well as food enthusiasts. Watching the different choices in food, this social web engineering model delves to equanimity and attention to seek the right locality of certain cuisines. In addition, the model takes the opportunities for the effort to share fine speech with ratings and photos uploaded to assist in deciding dining choices. The peaceful location, especially with good healthy promotions, would help the culinary businesses to start the presentation of menu and events from the beginning. This research moves on to unlock the healthy choice filled with exploratory hats to make every day fantastic learning adventures to ready and learn social engineering guidance. In conclusion, the mobile mechanism can help to present important pieces of advice to learn and practice useful culinary informatics.

Diabetes Risk Prediction using Shapley Additive Explanations for Feature Engineering

2025-01-04T14:22:41+08:00

Diabetes is prevalent globally, expected to increase in the next few years. This includes people with different types of diabetes including type 1 diabetes and type 2 diabetes. There are several causes for the increase: dietary decisions and lack of exercise as the main ones. This global health challenge calls for effective prediction and early management of the disease. This research focuses on the decision tree algorithm utilization to predict the risk of diabetes and model interpretability with the integration of SHapley Additive exPlanations (SHAP) for feature engineering. Random forest and gradient boosting models were developed to identify the risk factors and compare the prediction with the decision tree model. The performance of these classifiers was evaluated using the metrics for accuracy, f1-score, precision, and recall. Understanding the features that drive predictions can enhance clinical decision-making as much as predictive accuracy. With the use of a comprehensive dataset having 520 instances with 17 features including the target output, the proposed decision tree model had an accuracy of 97%. The decision tree model’s categorical variables enable straightforward data visualization. The SHAP tool was applied to interpret the model’s prediction after developing the model. This is crucial for healthcare practitioners as it provides specific health metrics to identify high-risk diabetic patients. Preliminary results indicate that a combination of polyuria, polydipsia, and age are predictors of diabetes risk. This study highlights the benefits that the integration of SHAP and decision trees algorithm provides predictive capability and transparent model interpretability. It also contributes to the growing body of literature on machine learning in the healthcare industry. The results advocate for the application of this methodology in clinical settings for prediction fostering trust between the approach and practitioners and patients alike.

A Fundamental Framework for Analysis of Rainfall Prediction Features Significance

2025-01-21T05:49:11+08:00

Rainfall prediction efforts had been prevalent ever since the impact of climate change on occurrences of natural disasters globally. Implementation of machine and deep learning techniques on features that contribute to rainfall occurrences were conducted with aims of seeking greater prediction accuracy for rainfall occurrences with a lack of study for significance of features in rainfall occurrence prediction. This study presents a framework of rainfall prediction features' significance analysis in the case study of Peninsular Malaysia rainfall occurrences. Features investigated in this study consist of temperature, humidity and wind speed. The designed framework for the investigation includes phases of data collection, data preprocessing, integration of random forest (RF) for ensemble classification and feature importance (FI) for feature significance calculation and finally model evaluation based on the metrics of precision, recall, F1 score and receiver operating characteristic (ROC) curve. In the preliminary investigation, the prediction model demonstrated accuracy, precision, recall and F1-score of 80.65%, 80%, 81% and 0.80 respectively. Humidity was found to have highest significance to the model's predictive power as compared to temperature and wind speed. Rainfall occurrence correlation with lower temperature and higher humidity and vice versa was identified with further investigation of feature data distribution against rainfall occurrences.

Enhancing Citrus Plant Health through the Application of Image Processing Techniques for Disease Detection

2024-09-02T14:33:41+08:00

The foremost task in agriculture is the decisive identification of citrus plants and the timely identification of diseases in the plants with the aim of improving the quality of crops and the yield. In this work, a machine learning algorithm focuses on image processing of citrus to solve issues that are significant and cause concern in agriculture. This work focus on the machine learning models like VGG 19 and VGG 16. In addition, dataset curation, data augmentation and various other methods were employed. The dataset used in this research is a composed one which is recorded in a comprehensive manner including the data of both the affected and healthy pieces of citrus fruits. The ensemble model utilised here to ensure the improvement of trained datasets. Reviewing the research on machine learning models indicates a possibility for accurate classification of the fruits and disease detection models of the fruit. The three contenders performed admirably, with VGG 19 dominating with 95.5% accuracy. In second place was CNN with 93.4% and VGG 16 trailing at 91.2%. Such models are recognisable, because they perform well in agricultural environments, thanks to their precision, recall, and F1 scores, which are all balanced properly. The models’ capacity to lessen the number of false alarms and misses is further assessed with the use of confusion matrices, which are of utmost importance in disease control. New developments in early disease diagnosis and detection of citrus fruits in agriculture may greatly enhance the health and productivity of crops. This research can be critical in increasing agricultural productivity while ensuring the environmental sustainability and health of growers and citrus crops in the long run.

Performance Evaluation on COVID-19 Prediction using Machine Learning Models

2024-11-15T05:29:24+08:00

The COVID-19 pandemic has placed enormous strain on providing health care services internationally while reinforcing the argument for the need to strengthen forecasting techniques. Existing forecasting methods have drawbacks, especially in determining the long-term consequences of the pandemic and understanding its broad reach across various locations and populations. This project proposes an evaluation of machine learning (ML) models with the aim of improving predictions, particularly the accuracy in long-term forecasting, of subsequent trends of the COVID-19 pandemic. A systematic review highlights previous forecasting attempts as a reference for the approach. This project emphasizes extensive data collection, model formulation and testing to develop a strong prediction framework. The models considered for evaluation are Support Vector Regression (SVR), seasonal autoregressive integrated moving average (SARIMA), and artificial neural networks (ANN), which have overcome some of the deficiencies of epidemiological forecasting methods to date. The aim is to provide public health representatives with more rigorous forecasts, which could enhance planning and response measures and protect health and safety. Our findings show that the ANN model is superior, with high accuracy and comprehensive performance, confirming its broader use in various predictive applications. The Root Mean Square Error (RMSE) of prediction error was also relatively modest (R-square values were nearly 1).

Image-based Detection and Classification of Poultry Diseases from Chicken Droppings in Open House Poultry Farms

2025-02-15T06:50:33+08:00

Monitoring chicken health is essential for maintaining the production efficiency of poultry farms and meeting the demand for poultry products. Previous studies have explored various methods, including utilizing sound, behaviour, and the shape of the chickens, as well as the conditions of their droppings, to assess chicken health. In this research, we monitor chicken droppings as a reliable indicator of chicken health. We develop an automated system for detecting chicken droppings and identifying health conditions, specifically in open house poultry farms in Malaysia. Open poultry houses are the most common design in Malaysia due to their lower construction and maintenance costs, a more natural environment for the chickens, and greater space to roam. However, the design of open poultry houses, which utilizes evenly gapped wood slat flooring, compounds the problem of automatically distinguishing new droppings from dirty flooring. In our work, data consisting of chicken dropping images from a poultry farm in Malaysia were collected for analysis. We used the YOLOv5n algorithm for detecting chicken droppings and distinguishing between healthy and sick chickens based on observable features such as the colour and shape of their droppings. Our proposed architecture, which used the YOLOv5n algorithm, can accurately detect chicken droppings and classify them into three health classes (coccidiosis, healthy, and other unhealthy), with an accuracy rate of up to 94.9%. By leveraging advanced computer vision techniques, poultry farmers can benefit from timely and accurate health assessments, leading to improved productivity and animal welfare in open house poultry farming systems.

Machine Learning Model for Assessing Human Well-being Using Brain Wave Activities

2024-09-28T16:13:37+08:00

This study presents a novel machine learning approach to assess human well-being through the analysis of brain wave activities. We developed a Random Forest classifier to categorize brain wave patterns into three states of well-being: good, normal, and bad. Using synthetic data simulating electroencephalography (EEG) readings, our model achieved an overall accuracy of 96.17%. The feature importance analysis revealed that alpha waves (34%) and beta waves (29%) were the most significant predictors of well-being states, which aligns with existing neuroscientific literature linking alpha activity to relaxation and beta activity to cognitive engagement. The confusion matrix demonstrated the model's particular strength in distinguishing between optimal and suboptimal well-being states, with no misclassifications between these extremes. ROC curve analysis further confirmed excellent discriminative ability across all three classes, with AUC values ranging from 0.984 to 0.999. The study demonstrates the potential of machine learning in interpreting complex neurophysiological data for personalised health monitoring, potentially enabling real-time assessment and intervention strategies. While promising, the use of synthetic data necessitates further validation with real-world EEG recordings. This research contributes to the growing field of computational neuroscience and its applications in mental health and well-being assessment, potentially paving the way for more objective and personalised mental health interventions. Future directions include incorporating temporal dynamics, accounting for individual variability, and integrating multiple data sources for a more holistic approach to well-being assessment.

Cyber Security Threats of Using Generative Artificial Intelligence in Source Code Management

2025-02-27T20:50:59+08:00

Generative Artificial Intelligence (Generative AI) models are now broadly used for academic writing and software development for the sake of productivity and efficiency. Concerns on the impact of Artificial Intelligence (AI) tools on academic integrity and cybersecurity grow bigger with time. Generative AI is being used for code generation, editing, and review, raising ethical and security challenges. A big concern is the involuntary introduction of vulnerabilities into codebases. They can reproduce known bugs or malicious code that compromise software integrity because of the way models are trained: on large datasets. The tools may also pose additional security threats often encountered during software development. AI models trained on public data will generate code that resembles copyrighted content, creating ownership and legal grey areas. Use of AI to delegate coding increases potential adversarial attacks and model poisoning. Addressing these challenges would therefore call for a balanced approach towards AI integrating into software development. Secure coding practices, thorough testing, continuous monitoring, and collaboration between developers, security professionals, and AI researchers should be balanced. Strong governance, regular audits, transparency in AI development, and the embedding of ethical standards in AI usage will help in ensuring it is safe and effective. Generative AI should be seen as a tool to enhance, not replace, human expertise in software development. While automation can streamline workflows, developers must remain vigilant to detect and mitigate AI-induced vulnerabilities. A proactive approach that combines human oversight with AI-driven efficiency will be key to securing the future of software development.

A Conceptual Approach to Predicting Seismic Events and Flood Risks Using Convolutional Neural Networks

2025-02-02T10:10:33+08:00

This paper explores the application of convolutional neural networks (CNNs) in predictive modelling for seismic events and flood risks, with a particular focus on forecasting extreme quantile events that exceed historical data limits. Traditional risk assessment methods often struggle to estimate such extremes, highlighting the need for more advanced predictive models capable of handling rare but high-impact events. This research enhances CNN architecture to improve accuracy in high quantile predictions by integrating multi-source spatiotemporal data, addressing a critical research gap. The methodology involves incorporating diverse datasets, including geospatial, meteorological, and historical seismic or flood records, into CNN models to augment predictive capabilities. These models undergo systematic validation using historical events and real-world data to assess their reliability, robustness, and practical relevance. Furthermore, the study evaluates the potential of these advanced prediction models to inform disaster risk management and mitigation strategies. By leveraging deep learning techniques and optimizing CNN structures, this research aims to refine forecasting precision, supporting proactive disaster preparedness. The anticipated outcome is an improved predictive framework that enhances early warning systems, facilitates informed decision-making, and strengthens emergency response mechanisms. Ultimately, this study contributes to the broader goal of increasing resilience against natural disasters by equipping policymakers, emergency responders, and urban planners with more accurate and timely risk assessments.

Anomaly Detection in Network Traffic for Insider Threat Identification: A Comparative Study of Unsupervised and Supervised Machine Learning Approaches

2024-09-28T15:59:17+08:00

Insider threats pose a significant and growing risk to organizational cybersecurity, with recent studies indicating a 47% increase in insider incidents from 2018 to 2022. This paper presents a comparative analysis of unsupervised and supervised machine learning approaches for detecting potential insider threats through network traffic anomaly identification. We develop and evaluate an Isolation Forest (unsupervised) and a Random Forest (supervised) model, training them on a simulated dataset representing six months of network logs from a mid-sized company. Our study introduces a unique feature set combining traditional network metrics with temporal and behavioral indicators, enhancing the models' detection capabilities. Results show that the Random Forest classifier outperforms the Isolation Forest, with F1-scores of 0.6425 and 0.4624, respectively. However, the unsupervised approach shows promise in scenarios lacking labeled data. Key findings reveal that increased connection frequency and data transfer volume are critical indicators of potential threats, with temporal patterns also playing a significant role. This study provides valuable insights into the strengths and limitations of each approach, offering practical implications for real-world digital forensics investigations. We contribute to the field by proposing a hybrid approach that leverages the strengths of both methods, potentially improving the accuracy and adaptability of insider threat detection systems. These findings pave the way for more robust, context-aware cybersecurity measures in the digital age.

Analysis of Forensic Disk Imaging Tools for Data Acquisition and Preservation

2025-02-26T10:02:03+08:00

The identification, preservation, analysis, and presentation of electronic evidence to support legal or organizational inquiries constitute the discipline of digital forensics, which is crucial to contemporary investigations. A crucial component of forensic inquiry, disk imaging guarantees precision, dependability, and legal defensibility. To preserve the original evidence, disk imaging makes an identical, bit-by-bit duplicate of a digital storage device, capturing hidden data, deleted material, and active files. Given the critical role of disk imaging in forensic investigations, selecting the right tool is crucial for accuracy, efficiency, and compliance with forensic standards. This study assesses widely used tools, including AccessData FTK Imager, Guymager, X-Ways Forensics, OSForensics, and FTK Imager, to help researchers and industry professionals choose the most suitable option for their investigative needs. This research examines the usability, imaging speed, supported hashing techniques, supported output formats, and other aspects of each tool to assess their suitability for usage in various forensic scenarios. The shows that X-Ways Forensic is among the greatest imaging tools because of its wide range of supported operations, fast imaging speed, and format compatibility. The result of hash verification, perfectly matched with source data, again establishes the capability of AccessData FTK Imager, FTK Imager, Guymager, X-Ways Forensics, and OS Forensics to ensure forensic soundness. Its capability to generate a detailed report with comprehensive drive geometry and file segmentation establishes its applicability in forensic workflows. Besides, the time consumed for processing shows its applicability in time-critical investigations too.

Dynamic Job Recommendation by Profiling Undergraduates Academic Performances

2025-03-12T22:30:48+08:00

Job-seeking tasks are always challenging. Often, job recommendation systems require human intervention in the job-seeking process. Therefore, the study focuses on recommendation of most relevant job sectors and prioritizing companies based on a student’s profile. The objectives of this study are: (i) to identify important features that optimize job recommendation, (ii) to construct a predictive model that recommends most relevant job sectors, and (iii) to recommend companies by computing the similarity between student and job profiles. In this study, the dataset was collected from Graduate Tracer Study from a university. Additionally, a job dataset was collected to extend the training dataset. As a result, both students and job profiles are used in this study. To enhance the accuracy, several models have been utilized for classifying job sector. This includes both hierarchical and single level classification. In hierarchical classification, Random Forest and Categorical Boosting were utilized; while in single level classification, a total of 9 different machine learning models were utilized. To assess the model’s performance, the metrics such as accuracy, weighted precision, weighted recall, and weighted f1-socre, were utilized. The finding shows that Hierarchical Classification outperforms Single Level Classification, with evaluation metrics ranging from approximately 72% to 76%, whereas Single Level Classification achieved around 58% to 62%. In conclusion, the integration of BorutaShap with Bidirectional Encoder Representation Transformers with 12 transformed layers enhances the performance of Hierarchical Classification, with the highest evaluation metrics around 75%. To recommend companies, a predefined rule is utilized to filter relevant companies, then, the similarity of the companies is measured using Cosine Similarity after transforming both student and company information using Bidirectional Encoder Representation Transformers with 12 transformed layers.

Detecting Black Hole Attack using Support Vector Machine with XGBoosting in Mobile Ad-Hoc Networks

2025-03-17T11:37:09+08:00

Mobile Ad-Hoc Networks (MANET) is a type of ad-hoc networks which use less infrastructure, that means the nodes in this network forward the massages without the need of infrastructure such as routers, switches etc. One of the most used attacks that can affect MANET performance is the black hole attack. This attack leads to dropping the packets that means these packets will never arrive and it will decrease the delivery ratio for the packets. This attack is a real problem as the sender is not informed that the data has not reached the intended receiver. The main goal of this study is to propose a solution for detecting black hole attacks using Extreme Gradient Boosting (XGBoost) based on a Support Vector Machine (SVM), the system for detection seeks to examine network traffic and spot anomalies by examining node activities. Attacking nodes in black hole situations exhibit specific behavioural traits that set them apart from other nodes, the traffic under a black hole attack is created using an NS-2 simulator to test the effectiveness of this strategy, and the malicious node is then identified based on the classification of the traffic into malicious and non-malicious. The results of the proposed technique outperformed the existing machine learning techniques such as Neural Network (NN), SVM, k-Nearest Neighbors (KNN), Decision Tree (DT), Logistic Regression (LR), Random Forest (RF), AdaBoost-SVM in terms of accuracy score as it achieved 98.67% as well as other classification performance measures (Precision, Recall, and F-measure).

Performance Benchmarking: Pre-trained Models and Custom Convolutional Neural Networks in Deep Learning

2024-09-22T15:41:43+08:00

Recent advances in computer vision and deep learning, particularly Convolutional Neural Networks (CNNs), have significantly increased road safety. CNNs were used in this work to automatically detect and categorise traffic signs—a crucial task for autonomous vehicles (AVs) and advanced driver assistance systems (ADAS). These technologies' ability to accurately recognize traffic signs enables them to make informed decisions in real time, thereby elevating the standard for overall driving safety. The study uses a large, annotated dataset of images of traffic signs to train and assess the CNN model. We developed a model that can recognize a large number of traffic lights, even in challenging scenarios such as low light levels, adverse weather, or high traffic. CNN image processing enables the system to accurately recognize and categorize traffic signs. Real-time predictions made by the CNN model after training aid ADAS and autonomous vehicles in comprehending road conditions. Real-time recognition is essential for tasks like managing turns, stopping at red lights, and adhering to speed restrictions. The research also addresses real-world challenges to ensure the model performs effectively in light or weather changes. A thorough testing process validates the model's accuracy and reliability. Ultimately, this technology might significantly increase road safety by providing drivers with more precise information, improving ADAS and AV decision-making skills, and reducing the number of accidents caused by drivers misinterpreting traffic signals.

The Impact of Deep Learning in Brain Tumour Analysis

2024-12-16T10:58:42+08:00

The need for early and precise identification of abnormalities has made the detection and classification of brain tumours essential components of medical diagnosis. Because brain tumours are naturally complex and can have a wide range of sizes, shapes, and types, conventional diagnostic techniques like MRI interpretation and manual evaluations are difficult and time-consuming. Traditional methods frequently depend on human expertise, which is prone to errors, delays, and variability. Deep learning (DL) developments, on the other hand, have completely changed this field by providing increased automation, efficiency, and precision in tumour detection and classification because they can automatically extract pertinent features from MRI scans, Convolutional Neural Networks (CNNs) have shown impressive success in medical image analysis in recent years. CNNs improve the classification of tumour types like gliomas, meningiomas, and pituitary tumours by using multiple layers to find patterns in imaging data. Despite their efficiency, CNNs sometimes struggle with complex tumour patterns, requiring further enhancement in feature extraction. Vision Transformers (ViTs) have become a viable substitute to overcome this constraint. ViTs are especially good at identifying complex tumour structures because, in contrast to CNNs, they use self-attention mechanisms to capture global image dependencies. ViTs can perform better diagnostics by more thoroughly analysing entire MRI images. Additionally, hybrid methods that combine CNNs and ViTs have demonstrated better outcomes, taking advantage of both long-range spatial understanding (ViTs) and local feature extraction (CNNs). These developments allow for real-time medical applications, drastically improve diagnostic accuracy, and lower false positives. Neuro-oncology could undergo a revolution with the incorporation of DL models into clinical workflows, which would improve tumour detection's accuracy, speed, and accessibility. These techniques will be further developed in future studies, guaranteeing even higher accuracy and versatility in medical imaging.

A New Bw Index for Quantifying Scholars' Research Influence

2025-02-25T21:09:30+08:00

The growing importance of measuring and evaluating academic performance in academic hiring, promotions, funding allocation, and resource distribution has fueled the demand for better metrics. Traditional ranking indicators, such as publication count and citation-based indices, often fail to capture for the interdisciplinary influence and qualitative dimensions of research impact. These shortcomings highlight the need for more comprehensive evaluation metrics. The current study introduces a novel B_W Index, which integrates both quantitative and qualitative aspects of researcher contributions aiming to provide a more balanced and comprehensive evaluation of scholarly impact. For evaluating the effectiveness of proposed index, a comparative analysis was conducted on 200 researchers' profiles of Monash University Australia calculating both the h-index and the proposed B_W Index. The results of study indicate that researchers with identical h-index exhibit significant variation in B_W Index values ranging from 10 to 55, demonstrating its ability to distinguish research impact beyond citation counts. Furthermore, for researchers with an h-index of 20, the B_W Index ranges from 20 to 82, reflecting an increase in differentiation compared to traditional h index. These findings highlight the B_W Index as a more nuanced and equitable measure of academic influence, offering a refined approach to researcher evaluation and addresses the limitations of traditional metrics.

Facilitating Digital Experience Sharing Among Vehicles through Utilisation of Pre-existing Communication Infrastructure

2025-03-21T17:28:01+08:00

Vehicular communication applications are expanding quickly because of the approach of related technologies, such as vehicular cloud and the Internet of Vehicles (IoV). The combination of the Internet of Things (IoT) and smart transportation is the Internet of vehicles. Data related to infotainment, safety, and effectiveness with different vehicles and the Sustainable framework of vehicles can be exchanged. However, after the appearance of such empowering advancements, still a huge number of ideas need research. Data sharing (related trips and navigation) of new/old models and other new/old model vehicles with the owner's agreement should be taken care of. This paper proposes a novel technique which is a digital experience-sharing system. With the proposed system, vehicles can share their experience with different vehicles depending on the owner's authorizations. The technique of digital experience sharing will give vehicles the ability to share and reestablish past information and data (related trips and navigation). A traffic trace containing the information of the vehicle: longitude, latitude, trip information, time, and location. Open street map (OSM) and simulation of urban mobility (SUMO) tool have been used for the simulation of the proposed technique. Further, the structure of the message, for productive communication is provided with implementation details in this work. Additionally, the application is used in the vehicle, and the information related to the start, stay, and end points of the journey is stored on a cloud. After some time, the same place is visited by the same vehicle, and a notification about previous visit information is displayed by the application.

Exploring the Relationship Between Academic Achievement and ChatGPT Usage: A Survey of Higher Education Students in Malaysia

2025-03-17T11:36:25+08:00

Students' academic achievement is caused by several factors which include cognitive capability, study habits, instructor effectiveness, family support, socio-economic status, and access to modern learning technologies like ChatGPT. Since its introduction, studies have suggested that ChatGPT has significantly impacted students’ academic performance. This study aims to explore whether students who perform well academically in higher education are more likely to use ChatGPT to enhance their studies and to gather their opinions on this learning tool. A survey was conducted with forty-one students with excellent academic performance. The results of this study show that most students with high academic achievement view ChatGPT as a valuable learning tool, with 82.9% states that it helps them understand complex topics and 73.2% find it useful for assignments. 51.2% use ChatGPT for academic purposes and 39% use it for both academic and non-academic tasks which indicate its broader utility. 80.5% of students indicate it is beneficial for their studies even though only 49.8% trust its accuracy. 73.4% of students acknowledge the risk of misuse such as cheating even though 53.7% still believe current protections are sufficient. Finally, students suggest improvements in ChatGPT such as the ability to provide more accurate responses and handle complex academic queries. In summary, the study suggests that students who perform well academically do use ChatGPT to improve their studies and appreciate its benefits. However, they also raise ethical concerns regarding its potential of misuse. For educators, the outcomes of this study will be of great benefit to them as the results highlight the need for them to allow students to use ChatGPT to improve their academic performance and at the same time tackling potential unethical issues such as misuse and cheating.

Exploring User Perceptions of Security in Mobile Banking: A Study in Malaysia

2025-04-11T15:55:43+08:00

Mobile banking has become increasingly popular in Malaysia due to its convenience and efficiency. However, persistent security concerns such as vulnerabilities to cyber attacks, malware and phishing affect user trust and impede adoption. This study aims to explore how users perceive and experience the security features of mobile banking applications. In spite of the implementation of security controls, understanding user perceptions and experiences remains essential, as these are core factors which influence mobile adoption. Thus, this study investigates the user experience of mobile banking security features in Malaysia. Specifically, we focus on the users’ perceptions, satisfaction levels and concerns. We employed a hybrid approach consisting of online surveys in-person interviews, and online interview sessions for data collection. This study comprises a detailed assessment of user awareness and behavior regarding mobile banking security, insights into the impact of security features on the users’ trust and recommendations for enhancing the security and usability of mobile banking applications. Our findings revealed that there were significant variations in user experiences across diverse demographic groups, with younger users exhibiting higher security concerns. Furthermore, usability issues associated with confusing navigation and slow response times were reported, negatively affecting the overall user experience. From the analysis, we noticed that while users generally rated the overall Malaysian mobile banking security applications positively, they identified key areas for improvement. These include the need for enhanced access control authentication mechanisms for secured transactions and more intuitive security interfaces for user-friendly navigation and use. In summary, this study highlights that user perceptions and experiences are central to understanding mobile banking security concerns in Malaysia. Hence, user-centric security designs are desired for balanced protection with ease of use.

Generative AI-based Meal Recommender System

2025-04-28T19:09:41+08:00

Maintaining a balanced diet is essential for overall well-being, yet many individuals face challenges in meal planning due to time constraints, limited nutritional knowledge, and difficulty aligning meals with personal dietary needs. Traditional meal recommender systems often rely on predefined plans or collaborative filtering techniques, limiting their adaptability and personalization. This study presents a generative AI-based Meal Recommender System utilizing Variational Autoencoders (VAEs) to generate personalized and nutritionally balanced meal plans. The system processes user inputs, such as dietary preferences, nutritional goals, and ingredient availability, to provide tailored recommendations. VAEs effectively uncover hidden dietary patterns and nutritional relationships within complex data, facilitating relevant and personalized meal suggestions. The system is trained and evaluated using two integrated datasets: one containing detailed nutritional information for complete meal plans, including attributes such as calories, protein, fats, carbohydrates, and sodium, and another listing individual dishes along with their names and user ratings. The meal plan dataset connects multiple dishes into structured daily meal schedules, while the dish dataset provides popularity and quality insights through user feedback. Together, these datasets enable the generation of personalized and nutritionally optimized meal recommendations. Experimental evaluation indicates strong ranking performance with a Normalized Discounted Cumulative Gain (NDCG) score of 0.963. However, Root Mean Square Error (RMSE), Mean Squared Error (MSE), and Mean Absolute Error (MAE) scores of 47.77, 2282.32, and 36.28, respectively, highlight potential areas for improving nutritional accuracy. A practical comparison with existing meal recommendation applications demonstrates the VAE model’s advantages in terms of personalization, nutritional fine-tuning, and recommendation diversity. The research contributes to AI-driven nutrition planning, healthcare, and fitness, offering a scalable and intelligent solution for personalized dietary recommendations.

Exploring Big Data Management Approaches and Applications: A Case Study of Real-Time Data Analytics in Air Traffic Management

2025-03-15T17:22:48+08:00

The rapid proliferation of digital devices has generated vast amounts of data, presenting significant challenges in collection, processing, and analysis that traditional systems struggle to overcome. This study investigates big data management approaches, explicitly focusing on technologies capable of efficiently handling real-time data at scale. Within the context of Air Operations, we propose a Hadoop-based architecture designed to support the Observe-Orient-Decide-Act (OODA) loop and enhance air traffic management. By leveraging a distributed system deployed on a cloud-based platform, we demonstrate a cost-effective solution for optimised data processing and improved decision-making capabilities. Our analysis highlights the advantages of using Hadoop's distributed file system (HDFS) for managing both structured and unstructured data generated by various sensors and devices. Additionally, we explore the integration of real-time processing technologies, such as Apache Kafka and Spark, to facilitate timely insights essential for operational effectiveness. Cloud deployment not only enhances resource accessibility but also offers flexibility and scalability, which are crucial for adapting to the dynamic nature of defence operations. We also address critical considerations for security and compliance when handling sensitive military data in cloud environments and recommend strategies to mitigate potential risks. The study concludes with recommendations for addressing future technological needs in big data management, including the incorporation of machine learning for predictive analytics and improved data visualisation tools. By implementing our proposed architecture, the military/ civil aviation can enhance its operational efficiency and decision-making processes, positioning itself to meet future challenges in an increasingly data-driven environment.

Editorial: Augmented Intelligence for Enabling Knowledge-Driven Decision Making

2025-04-30T21:11:19+08:00

More flexible and cooperative decision-making processes are required as a result of society's digital transformation. The role of augmented intelligence, that is the synergistic fusion of artificial intelligence and human judgment in facilitating knowledge-driven strategies across various domains is examined in this thematic issue. The integration of business intelligence and software engineering, which forms the foundation for creating intelligent, scalable, and explicable systems, is essential to this investigation. The six chosen papers in this issue show how machine learning techniques can be used to mine and model both structured (such as health records indicators) and unstructured (such as product reviews, e-sports discourse, and social media text from X) data. Applications in political sentiment analysis, geopolitical opinion monitoring, risk communication related to weather, e-commerce consumer feedback, gaming community analytics, and mapping malnutrition for public health intervention are all covered in these papers. From explainability and interface design to data preprocessing and model deployment, software engineering is essential to coordinating these intelligent pipelines and guaranteeing that AI outputs are not only accurate but also practically sound. The pieces in this issue collectively demonstrate how Augmented Intelligence can transform decision-making in a rapidly changing digital society when enabled by domain-aware data pipelines and structured engineering frameworks.

K-Means Clustering Optimization of Toddler Malnutrition Status Using Elbow Method

2024-12-17T05:05:58+08:00

The problem of nutritional status is still a major challenge in the health sector in developing countries, including Indonesia. Malnutrition in toddlers can have serious long-term impacts on children's growth and development, including increased risk of disease, impaired cognitive function, and low productivity in the future. To overcome this problem, an in-depth analysis is needed to determine the distribution of nutritional status of toddlers in one of the provincial capitals in Indonesia, which can be used as a basis for planning more effective interventions. This study uses the K-Means algorithm to classify areas based on the prevalence of malnutrition in toddlers across all sub districts in the city. Determination of the optimal number of clusters was carried out using the Elbow method, which showed that the most appropriate clusters were two clusters. To assess the quality of the cluster, the Davies Bouldin Index (DBI) was used which produced a score of 0.361, while the Silhouette Score was 0.799, indicating that the cluster results were of high quality. The clustering results showed significant variations in the prevalence of malnutrition in various sub districts. Cluster 0 represents areas with low prevalence of malnutrition, comprising six sub districts, while Cluster 1 includes ten sub districts with high prevalence of malnutrition. By identifying these high-risk areas more clearly, health authorities and practitioners can develop more targeted and effective nutrition interventions. This research highlights the importance of data driven decision making in public health, supporting augmented intelligence in identifying and addressing nutritional problems in urban areas. The insights provided by this clustering approach contribute to more efficient and strategic health intervention planning.

Classification of Smartphone Product Reviews on E-Commerce using the Recurrent Neural Network (RNN) Method

2025-03-23T21:37:05+08:00

Understanding how consumers behave in e-commerce is essential for businesses, especially in today’s digital world where people rely heavily on online shopping platforms. A key part of this understanding comes from sentiment analysis, which looks at customer reviews to find out what buyers really think and feel about products. However, analysing these reviews is not always straightforward. Many people use informal language, slang, or mixed languages, which makes it hard for computers to interpret their opinions accurately. On top of that, there is often an imbalance in the types of data available, particularly in developing countries, where some opinions might be overrepresented while others are missing. In this study, we tackled these challenges by collecting a large number of smartphone reviews from a leading e-commerce site. We used a Recurrent Neural Network (RNN) with a bidirectional Long Short-Term Memory (LSTM) architecture, which is good at understanding the context and meaning in sequences of words. Our approach also involved optimizing the model with the Adam optimizer, using 100-dimensional word embeddings, and applying dropout regularisation to prevent overfitting. For comparison, we tested more traditional techniques, like Support Vector Machine (SVM) and Naïve Bayes, against our RNN model. By balancing the dataset with random oversampling, the RNN achieved an impressive accuracy of 95.13%, outperforming the baseline methods by 7–9%. Overall, our results highlight the potential of advanced neural network models in improving sentiment analysis for e-commerce platforms, especially in challenging environments. This research provides a practical foundation for future work in natural language processing and can help businesses better understand and respond to their customers’ needs.

Sentiment Analysis of the 2024 General Election Through Twitter using Long-Short-Term Memory Algorithm

2025-04-12T20:36:22+08:00

This study analyses sentiment related to the 2024 Indonesian Presidential Election using the Long Short-Term Memory (LSTM) algorithm. A total of 2,400 tweets in the Indonesian language were gathered, with approximately 400 tweets sampled per week. In the data preparation, lexicon-based sentiment tagging, oversampling for class balance, and the creation and training of an LSTM model are all included in the study approach. The built model consists of embedding layers, Conv1D, and two LSTM layers. The LSTM model was selected due to its ability to capture long-range contextual dependencies in sequential text data like tweets, facilitated by its gate mechanisms (input, forget, output) that regulate information flow. The model achieved 84.3% accuracy in classifying sentiments (positive, neutral, negative), demonstrating its potential for real-time public opinion monitoring. The results provide actionable insights for election organisers and political analysts. For further study, using a wider spectrum of data to supplement model performance will help development. Tweaking hyperparameters and playing with other architectural models like GRUs or Transformers could improve model accuracy. Improved sentiment tagging calls for a more thorough and relevant sentiment vocabulary. The proposed model can be further developed into a real-time sentiment analysis tool to provide insights into public opinion on elections and other concerns.

Sentiment Analysis and Topic Modelling on Twitter Related to Mobile Legends: Bang Bang Game Using Lexicon-Based, LDA, and SVM

2025-03-23T20:51:54+08:00

Mobile Legends: Bang Bang (MLBB) has become a significant phenomenon within the global e-sports landscape, attracting millions of active players and fans. This study presents a comprehensive sentiment analysis and topic modelling of MLBB-related discussions on Platform Twitter, combining a lexicon-based approach, Latent Dirichlet Allocation (LDA), and Support Vector Machine (SVM) classification within a unified analytical pipeline. A dataset of 4,313 tweets was analysed, revealing that 70.8% expressed neutral sentiment, suggesting that much of the community's communication is informational rather than emotionally charged. Positive sentiments were associated with game content updates and rewards, while negative sentiments focused on technical and competitive issues. The SVM model achieved a sentiment classification accuracy of 90.57%, and cluster classification reached 85.13%. These findings offer valuable insights into how players engage with the game and reflect the underlying sentiments that influence the perception of gameplay and system updates. Furthermore, the predominance of neutral sentiment suggests opportunities for developers and content creators to enhance emotional resonance and community interaction through more engaging content and responsive design. The effectiveness of the combined methodology demonstrates the potential of integrating lexicon-based techniques with machine learning and topic modelling in analysing social media discourse within gaming communities. Future research is recommended to adopt advanced deep learning techniques, develop domain-specific sentiment lexicons, conduct multilingual sentiment analysis, and perform temporal tracking of community sentiment over time, enabling more dynamic and inclusive assessments of user experience and satisfaction.

Sentiment Analysis of the Israel-Palestine Conflict on X: Insights from the Indonesian Perspective using a Long Short-Term Memory Algorithm

2025-03-28T17:17:52+08:00

The Israel-Palestine conflict which has persisted for decades drives mounting global interest that consequently influences public opinion worldwide. This article examines the sentiment analysis of X (Twitter) data pertaining to the conflict using the Long Short-Term Memory (LSTM) model. This study presents public reactions through an analysis of 1,700 tweets collected between May and July 2023 which encapsulate key recent developments. In this study, several steps were conducted, namely 1) crawling process to get raw data; 2) preprocessing: cleansing, case folding, tokenization, stop word removal, and stemming; 3) modelling and validation using the LSTM model; 4) model evaluation based on performance metrics to evaluate the ability of the classification model to distinguish between classes; 5) visualization of experimental results. The LSTM model is a modification of the recurrent neural network (RNN). The LSTM model has many advantages, including being able to remember a collection of information that has been stored for a long period of time, being able to delete information that is no longer relevant, and being more efficient in processing, predicting, and classifying data based on a certain time sequence. Another advantage is that LSTM's ability to identify temporal dependencies and contextual interactions in sequential data makes it suitable for social media text analysis. The model demonstrated success in sentiment classification on geopolitical topics with an impressive accuracy rate of 91%. The findings demonstrate deep learning's potential applications for sentiment analysis and offer insights into public opinion dynamics during times of international crises.

Aspects-Based Sentiment Analysis of Extreme Weather on Twitter Using Long Short-Term Memory

2025-03-23T11:17:57+08:00

This study presents an aspect-based sentiment analysis of tweets related to extreme weather events in Indonesia, utilizing the Long Short-Term Memory (LSTM) model. The dataset was obtained through a Twitter crawling process, followed by a series of preprocessing steps including data cleaning, stop word removal, normalization, tokenization, and stemming. The three primary areas of emphasis in the study were kinds of bad weather forecasts, and the government or society reactions. Using a lexicon-based technique, sentiment labelling generated three groups: positive, neutral, and negative. A random oversampling method was employed to address the data imbalance. The model using the LSTM algorithm was trained individually for aspect and sentiment classification tasks, so reaching high accuracies of 98.94% and 97.53%, respectively. The results indicate that the model effectively categorises talk on extreme weather and the opinions of the public. A word cloud visual representation was additionally created to show frequently occurring terms in the dataset, thereby offering insights into current themes and sentiment expressions. This work provides valuable input for government agencies and legislators in developing communication and disaster response plans, thereby serving to better understand the public's view on climate-related events. Future work could involve improving techniques for preprocessing and using larger, wider-ranging datasets for improving the model's robustness and generalisation.