cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Furizal
Contact Email
furizal.id@gmail.com
Phone
-
Journal Mail Official
sjer.editor@gmail.com
Editorial Address
Jl. Poros Seroja, Kesra, Kepenuhan Barat Sei Rokan Jaya, Kec. Kepenuhan, Kab. Rokan Hulu, Riau
Location
Kab. rokan hulu,
Riau
INDONESIA
Scientific Journal of Computer Science
Published by PT. Teknologi Futuristik Indonesia
ISSN : -     EISSN : 31103170     DOI : https://doi.org/10.64539/sjcs
Core Subject : Science,
Computer Science & IT
The Scientific Journal of Computer Science (SJCS) (e-ISSN: 3110-3170) is a peer-reviewed and open-access scientific journal, managed and published by PT. Teknologi Futuristik Indonesia in collaboration with Universitas Qamarul Huda Badaruddin Bagu and Peneliti Teknologi Teknik Indonesia. The SJCS dedicated to publishing high-quality research across all areas of computer science, with a particular focus on emerging technologies that are shaping the future of computing. SJCS invites original research, review papers, and studies that involve practical applications, simulations, and theoretical advancements. The journal scope includes, but is not limited to: Artificial Intelligence and Machine Learning Data Science and Big Data Cybersecurity and Cryptography Cloud Computing and Distributed Systems Software Engineering Human-Computer Interaction Computer Vision and Natural Language Processing Internet of Things (IoT) Blockchain Technologies Robotics and Automation Computational Biology and Bioinformatics All fields related to computer science SJCS aims to advance the development of innovative computing systems that contribute to technological progress across industries.
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 28 Documents
 1   2   3 
An Intelligent Conversational Agent for Flood Risk Communication in a Flood-Prone Region of Nigeria Ebipamobonumugha, Willie; Onwudebelu, Ugochukwu; Kokogbiya, Efe Darel; Ogoja, Justina
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.411

Abstract

Flooding remains one of the most devastating natural hazards in developing countries, with significant impacts on human lives, infrastructure, and livelihoods. In Nigeria, particularly in Bayelsa State, recurrent flooding events highlight the need for effective and accessible flood risk communication systems. However, existing approaches largely rely on static and non-interactive dissemination channels, limiting timely public engagement and response. This study addresses this gap by designing and implementing a conversational agent capable of providing real-time responses to frequently asked flood-related questions. The proposed system adopts a rule-based conversational framework supported by natural language preprocessing techniques, including tokenization and normalization, for query interpretation. A structured knowledge base containing flood preparedness and response information was developed for the study area. The system was evaluated using a set of 120 representative flood-related queries derived from domain-specific scenarios. Experimental results show that the chatbot achieved a response accuracy of 87.5% and a successful query handling rate of 90.8%. These findings demonstrate the feasibility of conversational agents as effective tools for enhancing flood risk communication and public awareness. The study contributes to the integration of artificial intelligence-driven solutions into disaster risk management and highlights the potential of chatbot systems in improving access to critical environmental information in resource-constrained settings.
Machine Learning-Based Diabetes Classification Using Vital Signs and Clinical Information from the MIMIC-IV Dataset Huynh, Huy; Cao, Thanh; Tran, Hai
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.439

Abstract

Diagnosing diabetes based on clinical data is very important because the number of people with diabetes is growing around the world. The main focus of this study is on using machine learning models to figure out what kind of sickness someone has from a variety of clinical data. The MIMIC-IV dataset was used, which has both structured and unstructured data. The structured data includes vital signs, demographics, and lab tests. The unstructured data includes medical notes, major complaints, and a list of medications. Logistic Regression, Random Forest, Gradient Boosting, Support Vector Machine, and XGBoost were some of the models that were tested. Accuracy, Precision, Recall, F1-score, and AUC-ROC were used to measure how well the models worked. When random text data was added to the experiments, the results showed a big improvement in performance: the accuracy increased from approximately 68% to up to 87% across models. The best-performing models achieved AUC-ROC values above 0.95, with Random Forest and XGBoost showing the strongest performance. This shows that semantic mining from clinical notes is a key part of making medical decision support systems more reliable.
Predictive Analytics Model for Adaptive Teaching in Open and Distance Learning Institutions: Machine Learning Approach Adayilo, Danladi Moses; Oyefolahan, Ishaq Oyebisi; Ndunagu, Juliana Ngozi; Anekwe, Nwando; Malcalm, Ebenezer; Twabu, Khanyisile
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.451

Abstract

The study investigates the application of predictive analytics model in adaptive teaching within Open and Distance Learning (ODL) institutions. The aim of the study lies in addressing the ongoing challenges of high dropout rates and low student engagement, particularly in developing countries. The research gap is the underutilisation of predictive analytics to personalise interventions and enhance learning outcomes in ODL environments. The study employs mixed-method research design including machine learning algorithms with Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and XGBoost, in predicting students at risk of academic failure and providing personalised interventions. A dataset of 5,000 students from the National Open University of Nigeria was used to trained and test the model. Model validation metrics used includes: accuracy, precision, recall, F1-score, ROC-AUC, and PR-AUC. More so, (n=1050) participants took part in the experimental and control group including semi-interview, enabling real world application of predictive model. Key findings indicated that Random Forest had the highest ROC-AUC (98.38%), followed by XGBoost (97.76%). Nevertheless, Logistic Regression and SVM outperformed the others in accuracy (97.43%), precision (97.65%), recall (95.95%), and F1-score (96.79%). These results show that adaptive teaching, supported by predictive analytics, is associated with improved student engagement and contributes to reducing dropout rates. The challenges such as data quality, privacy, trust and algorithms bias should be addressed. The study suggest that predictive analytics is capable of transforming teaching methods in ODL institutions, improve personalised and effective learning. Future study should focus on model optimisation and integration with other educational technologies.
Hybrid Neuro-Fuzzy Deep Learning with Genetic Optimization for Explainable Stock Price Forecasting in Emerging Markets Adewale, Olumide Sunday; Ibam, Emmanuel Onwuka; Oluwagbemi, Johnson Bisi
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.377

Abstract

Precise stock price forecasting is vital for economic stability and capital allocation, yet it remains a tenacious challenge in emerging economies due to the inherent uncertainty and non-linearity of financial time series. Despite advances in deep learning, existing models often lack linguistic interpretability, fail to adapt to rapid market shifts, or exhibit look-ahead bias due to static validation splits. Moreover, empirical research focused on African financial systems, such as the Nigerian market, remains sparse, limiting the practical utility of conventional black-box architectures. This study proposes a Hybrid Neuro-Fuzzy and Deep Learning (HNFDL) framework that integrates fuzzy inference systems with Long Short-Term Memory (LSTM) networks and Genetic Algorithms (GA). The objective is to unify semantic reasoning with temporal learning to improve forecasting accuracy while maintaining high model transparency through explainable AI (XAI). Empirical validation using data from the Nigerian Exchange Group (NGX) (Dangote Cement, Zenith Bank, and the NSE All-Share Index) shows that the HNFDL model achieved a directional accuracy of 68.4% and a Mean Absolute Percentage Error (MAPE) as low as 4.36%. An ablation study confirmed that GA-driven optimization reduced the Root Mean Square Error (RMSE) by 8.4%, while the Diebold-Mariano test () statistically confirmed the model's superiority over standalone LSTM and fuzzy baselines. These results demonstrate that combining explainable fuzzy reasoning with adaptive deep neural architectures significantly enhances decision-making confidence. The framework provides a robust, statistically validated decision-support tool for investors and policy makers operating within volatile, information-asymmetric financial environments.
Graph Neural Networks (GNN) and Long Short-Term Memory (LSTM) for Forecasting Learner Attrition: A Systematic Review Otuya, Chinedu Cory; Obiniyi, Afolayan Ayodele; Igwe, Joseph Sunday
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.453

Abstract

The issue of learner attrition is a long-standing problem in Open and Distance Learning (ODL) settings where the lack of physical interaction and flexibility exacerbates the risk of disengagement. The use of Deep Learning (DL) techniques for forecasting complexity of behavioural and relational patterns of educational data has grown in usage. While Artificial Intelligence, DL in particular offers superior accuracy in forecasting attrition, the selection of appropriate techniques that addresses temporal sequence and relational patterns remains a critical gap due to inductive biases of ODL settings. This paper performs a systematic review based on the Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA) tool in order to synthesize the current body of knowledge regarding the use of LSTM and GNN in forecasting attrition. The peer-reviewed articles were located in major digital databases and filtered based on predetermined inclusion and exclusion criteria. The review evaluated model archetypes, data properties, metrics of evaluation, and performance results. Results showed that LSTM models were more useful in learning temporal patterns of engagement, whereas GNN models were efficient at learning relational and social learning patterns. Nevertheless, differences in datasets, validation procedures and evaluation metrices made it difficult to directly compare the results. The study identified methodological gaps of single models and recommended the use of hybrid methods for increased accuracy. The review gave consolidated information that direct researchers and institutions in the selection of suitable hybrid deep learning model in forecasting learner attrition.
Software Ecosystem Architectural Challenges and Mitigation Strategies: A Systematic Literature Review Ur Rahman, Inayat; Ur Rahman, Atta; Shahzad, Sara; Ur Rahman, Sajid
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.456

Abstract

Software ecosystems (SECO) play a crucial role in modern software development by enabling accelerated innovation, collaboration among multiple stakeholders, and efficient utilization of shared resources and technologies. However, achieving these benefits requires robust, adaptable, and well-structured architectural design and management. Despite their importance, SECO architectures face several critical challenges, including interface instability, security vulnerabilities, scalability limitations, governance complexity, sustainability concerns, and evolving ecosystem dynamics. Although prior studies have explored individual aspects of SECO, there is a clear research gap in providing a comprehensive and systematic synthesis of architectural challenges and their corresponding mitigation strategies. In particular, no systematic literature review (SLR) has thoroughly examined these issues in an integrated manner. To address this gap, this study aims to systematically identify, categorize, and analyze architectural challenges in SECO and evaluate existing mitigation techniques. A structured SLR methodology is employed to collect, assess, and synthesize relevant literature, leading to the development of a conceptual framework that organizes both challenges and solutions. The findings reveal that key mitigation strategies—such as modularization, variability management, custom design approaches, and sandboxing—can significantly improve architectural stability, scalability, and sustainability. These results provide valuable insights for both researchers and practitioners by offering a consolidated understanding of SECO architectural issues and practical guidance for designing more resilient and sustainable software ecosystems.
A Structured Survey of Attention Mechanisms in Audio-Visual Fusion: Architectures, Challenges, and Evaluation Frameworks Donatus, Rexcharles Enyinna; Awodele, Oludele; Oguike, Osondu Everestus; Sambo-Magaji, Amina
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.438

Abstract

Audio-visual fusion plays an important role in multimodal artificial intelligence, particularly in applications such as speech processing, emotion recognition, and video understanding, where information from sound and vision improves performance and contextual understanding. Recent developments are driven by attention mechanisms and transformer-based models, which enable more flexible and context-aware interaction within and across modalities compared to conventional fusion approaches. Despite these advances, challenges remain, including sensitivity to noisy or missing modalities, modality imbalance, limited interpretability, and high computational cost. This paper presents a structured survey of attention mechanisms in audio-visual fusion, with emphasis on architectural design and evaluation practices across multiple application domains. A structured survey methodology inspired by PRISMA principles is used to identify and select relevant studies, followed by comparative analysis of model architectures, training strategies, and evaluation methods. The findings show that transformer-based and attention-centered architectures have become increasingly prominent and achieve strong performance across tasks. However, these approaches involve trade-offs between robustness, interpretability, and computational efficiency, and remain sensitive to noise and modality imbalance. Evaluation practices are also inconsistent, with limited use of standardized and robustness-focused metrics. The survey provides an attention-centered taxonomy of audio-visual fusion methods and synthesizes current approaches and evaluation strategies. It identifies key challenges and outlines directions for improving robustness, interpretability, and efficiency in practical deployment.
Deep Learning for Venomous and Non-Venomous Snakes Classification Lidani, Yakubu Abubakar; Yola, Abdullahi Musa; Tasiu, Abu; Sani, Nura Muhammad; Gidado, Sulaiman Muhammad
Scientific Journal of Computer Science Vol. 2 No. 2 (2026): December Article in Process
Publisher : PT. Teknologi Futuristik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.64539/sjcs.v2i2.2026.463

Abstract

Snakes are a major health threat in various communities, specifically where human and snake encounters are frequent. When a snake is not identified correctly, healthcare providers often administer the wrong treatment, this can worsen patient recovery outcomes or even prove fatal to the victim. Therefore, a fast, proper and accurate distinction between venomous and nonvenomous snakes is vital for proper antivenom administration. This study proposes a hybrid deep learning system combining a CNN and an LSTM model for snake image classification through feature extraction from visual data. The CNN extracts key spatial features such as colour and scale patterns, texture, and body shape, whereas the LSTM captures sequential dependencies across these features, by helping distinguish visual similarity amongst the species. The model was trained and evaluated on a dataset of 6,798 snake images from diverse sources. The system achieved a performance of 97% accuracy, 97% precision, 96% recall, an F1-score of 97%, and a ROC-AUC of 0.97. These results demonstrate that integrating CNN and LSTM is moderately effective for snake classification. The proposed system has practical applications in the area of emergency healthcare, wildlife management, as well as mobile based identification tool. With 97% accuracy, this model can improve emergency responders first aid, enhance a safer treatment administration and help make safer decisions on the use of antivenom, by reducing treatment delays and improving patient survival prognosis. This model has the potential to save lives and minimize the consequences of snakebite envenoming.

Page 3 of 3 | Total Record : 28

 1   2   3 

Filter by Year

2025 2026


Reset
Filter By Issues
All Issue Vol. 2 No. 2 (2026): December Article in Process Vol. 2 No. 1 (2026): June Vol. 1 No. 2 (2025): December Vol. 1 No. 1 (2025): June