Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Journal of Embedded Systems, Security and Intelligent Systems
Afrizal Doewes
Universitas Sebelas Maret
Computer Science & IT

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Forecasting User Perception of Steam Game Reviews across Multiple Genres: A BERT Sentiment-Topic Framework with Topic Attribution Muhammad Pramudito Priambodo; Afrizal Doewes; Arif Rohmadi
Journal of Embedded Systems, Security and Intelligent Systems Vol 7 No 3 (2026): September 2026
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v7i3.13290

Abstract

Purpose – This study develops an interpretable framework for monitoring and forecasting user perception of Steam games by integrating sentiment analysis, topic modelling, short-horizon forecasting, and topic-level attribution. Methods – English-language Steam reviews from fifteen games across five genres were analysed over a twelve-month period. BERT was used to generate review-level sentiment ratings, which were aggregated into a weekly User Perception Score (UPS). BERTopic identified discussion themes, while an expanding-window moving-average model forecast UPS over a four-week horizon. Forecast performance was evaluated through walk-forward cross-validation against naive and damped linear-regression baselines, and topic-level attribution was used to explain recent changes in perception. Findings – The sentiment model showed strong agreement with Steam’s binary voting signal, while the expanding-window forecaster generally produced lower prediction error than the comparison baselines. Genre-level patterns indicated more favourable perception for Simulation, Role-Playing, and Action-Adventure titles, whereas First-Person Shooter and Strategy titles showed more mixed perception. Topic attribution further revealed that changes in both topic sentiment and topic prevalence contributed to shifts in weekly UPS. Research Implications – The framework provides developers and publishers with an interpretable monitoring approach for identifying perception trends and the discussion themes associated with them. Originality – The study combines weekly perception forecasting with a decomposition of topic-level contribution, enabling dynamic and interpretable analysis beyond static sentiment or topic summaries.
Benchmarking Indonesian Transformer Models and Explainable AI for Disaster-Related Sentiment Analysis Muhammad Saifuddin Eka Nugraha; Afrizal Doewes; Herdito Ibnu Dewangkoro
Journal of Embedded Systems, Security and Intelligent Systems Vol 7 No 3 (2026): September 2026
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v7i3.13295

Abstract

Purpose – This study evaluates the comparative performance of Indonesian transformer models for disaster-related sentiment analysis and examines the faithfulness of explainable artificial intelligence methods applied to the best-performing model. Methods – YouTube comments related to the 2025 Sumatra flood were processed using a hybrid labeling approach combining automatic classification and expert annotation. After preprocessing and class balancing, IndoBERT, IndoBERTweet, and IndoRoBERTa were fine-tuned using Optuna-based hyperparameter optimization. Model performance was assessed using accuracy, precision, recall, and F1-score. Integrated Gradients (IG), Local Interpretable Model-Agnostic Explanations (LIME), and SHapley Additive exPlanations (SHAP) were subsequently evaluated using the Area Under the Threshold-Performance Curve (AUC-TP) to quantify explanation faithfulness. Findings – IndoRoBERTa achieved the strongest overall classification performance among the evaluated models. Faithfulness analysis showed that IG provided the strongest overall explanation performance and performed particularly well for negative and positive sentiment, whereas LIME showed better performance for neutral sentiment. SHAP produced comparatively weaker faithfulness under the applied evaluation protocol. Research Implications – The findings demonstrate the potential of Indonesian transformer models and quantitative XAI evaluation for analyzing disaster-related social media discourse. However, the results should be interpreted cautiously because automatic labeling, confidence-based undersampling, and the absence of inferential significance testing may affect generalizability. Originality – This study integrates comparative benchmarking of Indonesian transformer models with systematic deletion-based faithfulness evaluation of multiple XAI methods, extending explainable sentiment analysis beyond predominantly qualitative interpretation.
Co-Authors Arif Rohmadi Herdito Ibnu Dewangkoro Muhammad Pramudito Priambodo Muhammad Saifuddin Eka Nugraha