Article Per Year (5 Year)
p-Index From 2021 - 2026
0.702
P-Index
This Author published in this journals
All Journal Brilliance: Research of Artificial Intelligence
Yudi Wibisono
Universitas Pendidikan Indonesia, Indonesia
Decision Sciences, Operations Research & Management Mathematics Other

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
Intelligent News Aggregation System with Automatic Classification, Clustering, and Summarization Ihsan Ghozi Zulfikar; Yudi Wibisono; Asep Wahyudin
Brilliance: Research of Artificial Intelligence Vol. 5 No. 2 (2025): Brilliance: Research of Artificial Intelligence, Article Research November 2025
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v5i2.6712

Abstract

The rapid growth of online news content has made it increasingly difficult for users to access relevant information efficiently. This study presents the development of an intelligent web-based news aggregation system that performs automatic classification, clustering, and summarization of Indonesian-language news articles. The system aims to enhance the news reading experience by organizing articles by category and topic, and by providing concise summaries. The system was built using the ADDIE development model, with each AI component trained and evaluated separately. News classification is handled by a BLSTM-2DCNN model trained on the IndoSum dataset, achieving 86% accuracy and an F1-score of 0.85. This model was also applied to classify 37,187 real-world articles scraped from Kompas and TribunNews during June 2025. Topic clustering is performed using K-means with entropy-weighted Bag-of-Words features over 5-day sliding windows. The clustering quality, evaluated using the Calinski-Harabasz Index, ranged from 5.21 to 525.44 with an average of 80.53, indicating varying cluster cohesion. For summarization, a fine-tuned BART model was used to summarize the article closest to each cluster’s centroid. The model achieved ROUGE scores of 0.6389 (ROUGE-1), 0.5458 (ROUGE-2), and 0.6017 (ROUGE-L). The integrated system automatically scrapes news, classifies and clusters articles, and displays generated summaries through a user-friendly web interface. The results show that combining deep learning and natural language processing offers an effective approach for intelligent news aggregation, helping users consume news faster and more meaningfully.
Development of an Academic Services Chatbot Based on Retrieval-Augmented Generation (RAG) Mohammad Labib Husain; Yudi Wibisono; Ani Anisyah
Brilliance: Research of Artificial Intelligence Vol. 5 No. 2 (2025): Brilliance: Research of Artificial Intelligence, Article Research November 2025
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v5i2.6719

Abstract

Higher education institutions struggle to provide accurate and accessible academic information. Traditional chatbots are often limited in capability, while standard Large Language Models (LLMs) pose a significant risk of factual "hallucinations," rendering them unsuitable for official university use where trustworthiness is paramount. This study aims to increase the accessibility and effectiveness of academic services by developing a trustworthy chatbot. The primary objective is to implement the Retrieval-Augmented Generation framework to create a reliable AI assistant that is factually grounded in a verified, domain-specific knowledge base. A knowledge base was constructed from official FPMIPA UPI documents and structured using hierarchical chunking. The system employs a multi-stage RAG pipeline featuring query contextualization and reranking before generation with Gemini 2.5 Pro. Performance was evaluated using metrics from the RAGAS framework on a 100-question dataset categorized into factual, reasoning, and out-of-context queries. The evaluation revealed strong performance on factual queries, achieving a Faithfulness score of 0.9100. A significant performance decrease was observed for reasoning tasks, with Context Recall dropping to 0.5926. Crucially, the system successfully handled 81.25% of out-of-context questions by correctly refusing to answer, thereby effectively preventing hallucination. The RAG framework provides a viable and empirically-validated blueprint for creating a trustworthy conversational AI for higher education. The model proves to be an effective tool for factual information delivery and has strong potential to modernize how student support and academic services are delivered.
Development of an Automatic Summarization System based on Large Language Models for Annual Report Analysis Muhammad Rizki; Yudi Wibisono; Eddy Prasetyo Nugroho
Brilliance: Research of Artificial Intelligence Vol. 5 No. 2 (2025): Brilliance: Research of Artificial Intelligence, Article Research November 2025
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v5i2.6772

Abstract

The increasing interest in stock market investment in Indonesia has highlighted a significant challenge for retail investors: the difficulty of analyzing lengthy and complex corporate annual reports. These documents, essential for fundamental analysis, are often hundreds of pages long and contain detailed narrative sections that require considerable time and effort to comprehend. This research addresses this issue by developing an automatic summarization system using a Large Language Model (LLM) to generate concise and insightful summaries of such reports. The primary objective was to develop and evaluate an LLM-based system specifically adapted for the structure and content of annual reports. The method involved creating a tailored dataset comprising 2,008 narrative text excerpts and their corresponding manual summaries sourced from the annual reports of companies listed on the Indonesia Stock Exchange (IDX). The open-source Llama-3.2-3B-Instruct model was then fine-tuned using the Parameter-Efficient Fine-Tuning (PEFT) technique, specifically Low-Rank Adaptation (LoRA). The research results demonstrated a significant improvement in the model's performance after fine-tuning. Quantitative evaluation using ROUGE metrics showed a relative increase of 18.63% in ROUGE-1, 44.45% in ROUGE-2, and 33.83% in ROUGE-L compared to the base model. Qualitative analysis confirmed that the fine-tuned model was capable of generating informative and relevant summaries aligned with the context of annual report analysis. In conclusion, this study demonstrates that fine-tuning LLMs with document-specific data is an effective approach for specialized tasks such as annual report summarization.
Non-Playable Characters Based On Large Language Models For Role Playing Games (RPG) Ade Mulyana; Yudi Wibisono; Ani Anisyah
Brilliance: Research of Artificial Intelligence Vol. 5 No. 2 (2025): Brilliance: Research of Artificial Intelligence, Article Research November 2025
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v5i2.6779

Abstract

Interactive dialogue is a central element in role-playing games (RPG), particularly those that emphasize storytelling and immersion. This study explores the development of a dynamic Non-Playable Character (NPC) system using a Large Language Model (LLM) to simulate responsive conversations in a fictional world. The objective of this research is to design an NPC dialogue system that can maintain contextual consistency with the game’s lore while adapting to player input dynamically. The method used is engineering-based development, involving prompt engineering and a Retrieval-Augmented Generation (RAG) approach to embed narrative context into the LLM prompts. The system is implemented in a 2D RPG titled Kage no Meiyaku: Shinobi no Michi, where players interact with multiple NPCs whose responses evolve based on both pre-defined lore and game progression. Evaluation is conducted using a Likert scale across four dialogue quality dimensions: coherence, emotional engagement, narrative relevance, and persona consistency. The results show that the system generates engaging and contextually accurate responses, with average scores ranging from 4.0 to 4.5. Some limitations are identified, such as occasional misspellings and generic responses in ambiguous inputs. However, the approach demonstrates strong potential for AI-assisted storytelling in games. This research contributes to expanding LLM applications in interactive fiction and opens future work toward feature-rich RPG elements such as transactional systems, branching narratives, and real-time battle interactions.
Co-Authors Ade Mulyana Ani Anisyah Asep Wahyudin Eddy Prasetyo Nugroho Ihsan Ghozi Zulfikar Mohammad Labib Husain Muhammad Rizki