Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : bulletin of network engineer and informatics bufnets

Design of a Hybrid SVM Ensemble and Large Language Model Chatbot for Multi-Class Intent Classification in Clinic Information Services Vina Rahmadiany; Lindawati Lindawati; Aryanti Aryanti
Bulletin of Network Engineer and Informatics Vol. 4 No. 1 (2026): BUFNETS (Bulletin of Network Engineer and Informatics) April 2026
Publisher : PT. GWEX NET PUBLISHER

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59688/a3kdjj55

Abstract

Clinical information service chatbots require an accurate multiclass intent classification mechanism to handle informal language variations, medical abbreviations, service-related inquiries, and health complaints in Indonesian. This study aims to develop a hybrid chatbot architecture that integrates a Support Vector Machine ensemble for intent classification with Meta-Llama-3.1-8B-Instruct to generate relevant, natural, and context-aware responses. The dataset consisted of 3,293 utterance patterns across 66 intent classes. The proposed approach employed an 80:20 stratified split; semantic augmentation on the training data; preprocessing via abbreviation normalization, stopword removal, and Sastrawi stemming; TF-IDF feature extraction using unigram and bigram word n-grams and character n-grams; chi-squared feature selection; and hyperparameter optimization via grid search. The classification model was constructed by combining three calibrated LinearSVC classifiers via probability-based soft voting. Experimental results achieved an accuracy of 89.83%, a weighted F1-score of 89.90%, a kappa of 0.8966, a macro AUC of 0.9942, and an average response time of 61.87 ms. McNemar's test indicated statistically significant improvements over Complement Naïve Bayes, Logistic Regression, and Decision Tree, while no significant difference was observed compared with Single SVM. Therefore, the proposed architecture is effective at supporting multiclass intent classification and delivering fast, relevant chatbot responses for clinical information services.
Implementation of The Blake3 Algorithm for Detecting Changes and Ensuring The Integrity of Web-Based Digital Documents Putri Nabila; Aryanti Aryanti; Martinus Mujur Rose
Bulletin of Network Engineer and Informatics Vol. 4 No. 2 (2026): BUFNETS (Bulletin of Network Engineer and Informatics) October 2026
Publisher : PT. GWEX NET PUBLISHER

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59688/2hne3s53

Abstract

The increasing dependence on web-centric digital document management for archiving contracts, academic credentials, and administrative records has intensified anxieties regarding data manipulation and unauthorized alterations. Traditional cryptographic verification methods, including MD5 and SHA-1, are no longer viable due to demonstrated collision vulnerabilities. Meanwhile, more robust alternatives like SHA-256 and SHA-3 frequently encounter performance bottlenecks when processing large volumes of data in real-time web architectures. To address these limitations, this research integrates BLAKE3 a cutting-edge cryptographic hash function that pairs SHA-256-level security with superior throughput via a parallelized Merkle-tree design into a web-based platform dedicated to document integrity verification. Adopting a Research and Development (R&D) methodology framework alongside the Waterfall System Development Life Cycle (SDLC), the project progressed through requirements analysis, system architecture design, deployment, and evaluation phases. The prototype was built using a PHP/Node.js web stack backed by a relational database that logs the BLAKE3 hash of each uploaded asset, recalculating and cross-checking the signatures upon every retrieval query. System performance was assessed via black-box functionality assessments, avalanche effect trials, and hash execution time benchmarks contrasting BLAKE3 against MD5, SHA-256, and SHA-3. The empirical data demonstrates that the platform achieved a 100% success rate in identifying simulated file tampering. Furthermore, BLAKE3 outpaced SHA-256 by a factor of 3.7 to 8.1 depending on file dimensions, while yielding an optimal avalanche effect of 49.9%. These outcomes suggest that BLAKE3 serves as a highly scalable and dependable substitute for safeguarding digital assets within web information systems, particularly for platforms handling high-throughput document verification.