cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Adyanata Lubis
Contact Email
jmnr@rokania.ac.id
Phone
+628127651902
Journal Mail Official
jmnr@rokania.ac.id
Editorial Address
Jl. Raya Pasir Pengaraian,Km 15 Langkitin, Kec. Rambah Samo. Kab.Rokan Hulu
Location
Kab. rokan hulu,
Riau
INDONESIA
JOURNAL OF ICT APLICATIONS AND SYSTEM
Published by STKIP Rokania
ISSN : 28301404     EISSN : 2830098X     DOI : https://doi.org/10.56313/jictas
Core Subject : Science, Engineering,
Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering
The Journal of ICT Applications System is a scientific journal that presents original articles on computer science research. This journal is a means of publication and a place to share research and development work in the field of computers. Loading of articles in this journal is done through submit. Complete information for article loading and article writing instructions are available in each issue. Articles submitted will go through a selection process for bestari partners and/or editors. Journal of ICTAplication System is published 2 times a year, in June and December Journal of ICTAplication System Registered at PDII LIPI with Print ISSN number 2830-1404 and Online ISSN 2830-098X For practitioners, academics, teachers and students in the field of computer science who want articles on research results and ideas to be published in this journal via submit
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 43 Documents
 1   2   3   4   5 
Explainable Imbalance-Aware Spatiotemporal Learning for Traffic Accident Risk Prediction in Medan Metropolitan City Rusmin Saragih; Enda Ribka Meganta P; Theodora MV Nainggolan; Frans Ikorasaki; Fithry Tahel
Journal of ICT Applications System Vol 5 No 1 (2026): Journal of ICT Aplications and System
Publisher : Lembaga Penelitian dan Pengabdian Masyarakat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56313/jictas.v5i1.530

Abstract

Traffic accident prediction in rapidly urbanizing metropolitan regions remains a critical challenge due to the complex interplay of spatiotemporal dynamics, severe class imbalance, and the opacity of predictive models that limits actionable policy interpretation. Existing approaches tend to address these challenges in isolation—deploying graph neural networks without imbalance correction, or applying oversampling without incorporating spatial context—thereby falling short of the comprehensive decision-support capability demanded by intelligent transportation systems. This paper presents a novel integrated framework, designated SLT-SHAP, that systematically unifies spatiotemporal graph convolutional learning, Synthetic Minority Oversampling Technique (SMOTE) applied exclusively to the training partition, Long Short-Term Memory (LSTM) networks for sequential temporal dependency modeling, a Transformer encoder for long-range contextual attention across hourly traffic sequences, and SHapley Additive exPlanations (SHAP) for post-hoc model interpretability. The study employs a curated spatiotemporal dataset of 132,480 observations collected at hourly resolution across 48 administrative zones in Medan Metropolitan City, Indonesia, encompassing traffic, meteorological, infrastructural, and geospatial variables with an inherent accident class imbalance of 12.4%. Experimental results demonstrate that SLT-SHAP achieves an F1-score of 0.796, AUC-ROC of 0.963, AUPRC of 0.784, and Matthews Correlation Coefficient (MCC) of 0.783, surpassing all baseline and ablation variants. Ablation analysis confirms that each component—graph construction, SMOTE, LSTM, and Transformer—contributes independently to performance. SHAP analysis identifies congestion index, hour of day, and average speed as the three most influential predictors, with spatial heatmapping delineating persistent high-risk zones. The proposed framework offers a replicable and interpretable decision-support architecture for urban road safety analytics in the Indonesian and broader Southeast Asian metropolitan context.
TemporalXAI-Det: Temporal-Aware Explainable Detection of Multi-Model AI-Generated Academic Text via Continual Learning and Cross-Lingual Transfer Imeldawaty Gultom; Ratih Puspadini; Fauzi Erwis; Elyandri Prasiwiningrum; Ridwan
Journal of ICT Applications System Vol 5 No 1 (2026): Journal of ICT Aplications and System
Publisher : Lembaga Penelitian dan Pengabdian Masyarakat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56313/jictas.v5i1.531

Abstract

The proliferation of heterogeneous generative AI systems—including GPT-4o, Claude 3 Opus, Gemini 1.5 Pro, Mistral, and LLaMA-3—has produced a multi-source academic text landscape whose detection presents challenges qualitatively beyond those addressed by existing binary or single-source detection paradigms. Contemporary detectors are doubly compromised: first, by adversarial paraphrasing that disrupts surface-level distributional signatures; second, by temporal model drift, wherein new model generations evade detectors trained on earlier LLM families. This study introduces TemporalXAI-Det, a continual-learning explainable detection framework capable of (1) attributing academic text to one of five generative model families while simultaneously identifying human authorship, yielding a six-class taxonomy; (2) adapting to new LLM generations without catastrophic forgetting via Elastic Weight Consolidation (EWC) and experience replay; (3) transferring robustly across twelve academic languages through a Language-Adaptive Prefix Tuning (LAPT) mechanism applied to XLM-RoBERTa-XL; and (4) generating legally defensible per-instance explanations via Integrated Gradients (IG), SHAP, and counterfactual generation. A large-scale continual benchmark corpus (MTA-72K) comprising 72,000 samples across six source classes, four adversarial attack paradigms, and twelve languages is constructed and released. TemporalXAI-Det achieves a six-class macro F1-score of 0.941 on the clean test partition, 0.912 under combined adversarial conditions (performance degradation ? = 2.9 pp), and a mean cross-lingual F1 of 0.887 across all twelve evaluated languages. Continual learning experiments demonstrate that catastrophic forgetting is reduced by 78.4% relative to standard fine-tuning when new LLM families are introduced. These results establish new state-of-the-art benchmarks for multi-source, temporally robust, and multilingual AI-text detection in academic integrity contexts
A Hierarchical Buffer-Sizing Framework for Congestion Mitigation in Campus Area Networks: An Engineering-Theoretic Approach to the Internet Sluggishness Problem Monday Eyinagho; Olusegun Emoruoa
Journal of ICT Applications System Vol 5 No 1 (2026): Journal of ICT Aplications and System
Publisher : Lembaga Penelitian dan Pengabdian Masyarakat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56313/jictas.v5i1.532

Abstract

Campus Area Networks (CANs) in higher educational institutions worldwide have long suffered from Internet sluggishness—a persistent degradation in upload and download throughput that occurs at predictable temporal intervals. Prior investigations have narrowly attributed this phenomenon to insufficient bandwidth, recommending either bandwidth overprovisioning or policy-based management without resolving the problem. This paper advances a fundamentally different, engineering-theoretic explanation and a novel quantitative resolution: the sluggishness is primarily caused by the uniform deployment of identically buffered switches across all hierarchical layers of the network, which violates the traffic aggregation principle intrinsic to layered switched architectures. Using a physically installed university CAN at Afe Babalola University, Ado-Ekiti, Nigeria, we formally derive a Hierarchical Buffer-Sizing (HBS) framework grounded in graph-theoretic tree analysis. The proposed HBS framework yields per-switch buffer size specifications as a function of each switch's subtree cardinality within the network topology. Results show that the required buffer capacity for core-layer switches can be up to 14× greater than that of edge-layer leaf switches, a disparity completely absent in existing installations. Comparative simulation using NS-3 demonstrates that networks configured according to the HBS framework reduce average end-to-end queuing delay by 68.4% and packet drop rate by 73.1% relative to uniform-buffer baselines. The framework is analytically validated against both the small-buffer model of Appenzeller et al. [4] and the very-small-buffer model of Enachescu et al. [6], with all derived buffer values falling within theoretically acceptable bounds. This work provides, for the first time, a deterministic, topology-driven engineering methodology for CAN buffer provisioning that can be directly implemented by network engineers without traffic monitoring prerequisites

Page 5 of 5 | Total Record : 43

 1   2   3   4   5 

Filter by Year

2022 2026


Reset
Filter By Issues
All Issue Vol 5 No 1 (2026): Journal of ICT Aplications and System Vol 4 No 2 (2025): Journal of ICT Aplications and System Vol 4 No 1 (2025): Journal of ICT Aplications and System Vol 3 No 2 (2024): Journal of ICT Aplications and System Vol 3 No 1 (2024): Journal of ICT Aplications and System Vol 2 No 2 (2023): Journal of ICT Aplications and System Vol 2 No 1 (2023): Journal of ICT Aplications and System Vol 1 No 2 (2022): Journal of ICT Aplications and System Vol 1 No 1 (2022): Journal of ICT Aplications and System