Claim Missing Document
Check
Articles

Found 2 Documents
Search

A comparative study of GCN and GraphSAGE algorithms for structural anomaly detection in public procurement networks Danny Mario Tua Panjaitan; Eka Prima Ginting
Jurnal Sains, Teknologi & Komputer Vol. 3 No. 3 (2026): Jurnal Sains, Teknologi & Komputer (SAINTEK)
Publisher : Lembaga Riset Mutiara Akbar

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56495/saintek.v3i3.1677

Abstract

Bid rigging and procurement collusion may produce structural patterns in the relationships between companies and procurement opportunities. This study compares Graph Convolutional Network (GCN) and GraphSAGE for structural anomaly detection in a heterogeneous company–tender procurement network. The study uses Government Procurement via GeBIZ data from Singapore and introduces controlled synthetic anomalies because the dataset does not provide complete bidder-participation information or verified collusion labels. A leakage-aware temporal evaluation is applied together with multiple anomaly difficulties, random seeds, unseen anomaly subjects, and cross-pattern testing. The results show different empirical behavior between the two architectures. GraphSAGE achieves higher F1-Score, ROC-AUC, and PR-AUC in the multi-seed evaluation, while GCN achieves higher recall. GraphSAGE also performs better in the true inductive evaluation on unseen anomaly subjects. An Early Warning System based on GraphSAGE is further developed to rank companies according to their structural anomaly scores. The findings indicate that graph-based models can support structural anomaly screening, while synthetic anomaly results should not be interpreted as evidence of confirmed procurement collusion.
Comparative evaluation of deep learning models for aspect-level sentiment analysis using hybrid real and synthetic healthcare reviews Eka Prima Ginting; Danny Mario Tua Panjaitan
Jurnal Sains, Teknologi & Komputer Vol. 3 No. 2 (2026): Jurnal Sains, Teknologi & Komputer (SAINTEK)
Publisher : Lembaga Riset Mutiara Akbar

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56495/saintek.v3i2.1682

Abstract

Aspect-level sentiment analysis enables the identification of sentiment polarity toward specific aspects, providing more detailed information from healthcare reviews. This study aims to compare the performance of LSTM, BiLSTM, CNN-BiLSTM, and BERT for aspect-level sentiment classification. The dataset comprised a hybrid corpus of 5,000 reviews, including 2,000 real healthcare reviews and 3,000 AI-assisted synthetic reviews, covering five aspects: service quality, medical staff, facilities, waiting time, and cleanliness. The dataset was divided into 80% training data and 20% testing data. Model performance was evaluated using accuracy, precision, recall, and F1-score. The results showed that BERT achieved the highest performance, with 94.60% accuracy, 94.78% precision, 94.40% recall, and a 94.59% F1-score, followed by CNN-BiLSTM, BiLSTM, and LSTM. Medical staff achieved the highest aspect-level accuracy at 95.8%, whereas waiting time obtained the lowest at 92.7%. These findings indicate that BERT achieved the best classification performance among the four evaluated models under the experimental conditions of this study.