Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Blockchain Frontier Technology (BFRONT) Journal of Computer Science and Technology Application
Herman Herman
HKBP Nommensen University of Pematangsiantar
Computer Science & IT Control & Systems Engineering Education Library & Information Science Other

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

Found 2 Documents
Search

 1 
Self Learning Artificial Intelligence for Autonomous Threat Detection in Computer Networks Dwi Cahyono; Herman Herman; Ikyboy Van Versie
CORISINTA Vol 3 No 2 (2026): August
Publisher : Pandawan Sejahtera Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33050/tk5ypk40

Abstract

The rapid expansion of large-scale computer networks and the exponential growth of big data have significantly increased the complexity and frequency of cyber threats, rendering traditional signature-based security mechanisms inadequate for adaptive detection. This study aims to develop a self-learning AI model capable of autonomously identifying evolving attack patterns and anomalous behaviors in large-scale networks without relying exclusively on pre-labeled datasets. The proposed framework integrates deep neural architectures, incremental learning, and behavior-based traffic analysis to enable continuous adaptation to dynamic threat environments while ensuring computational efficiency and scalability. The model was trained and evaluated using realistic network traffic datasets simulating distributed attacks, zero-day exploits, and advanced persistent threats across heterogeneous environments. Experimental findings demonstrate that the self-learning approach enhances detection accuracy, reduces false positives, and accelerates response times compared to conventional intrusion detection systems. In addition, the combination of deep neural architectures with incremental learning and scalable data processing further strengthens model robustness and adaptability in complex and evolving networks. The results indicate that integrating adaptive AI into cybersecurity frameworks enhances proactive defense capabilities, improves resilience in large-scale computer networks, and provides a scalable, intelligent solution for next-generation threat detection systems. This study highlights the practical relevance of combining AI, big data analytics, and cybersecurity strategies to support intelligent, adaptive security solutions capable of addressing emerging threats, minimizing operational risks, and fostering robust network protection in increasingly complex digital infrastructures.
Utilizing Blockchain for Trustworthy and Transparent AI Decision Making Herman Herman; Rohim Rohim; Rizki Indrawan; Chua Toh Hua
Blockchain Frontier Technology Vol. 6 No. 1 (2026): Blockchain Frontier Technology
Publisher : IAIC Bangun Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34306/bfront.v6i1.1031

Abstract

The increasing adoption of AI in critical sectors such as healthcare, finance, transportation, and public services raises significant challenges related to transparency, accountability, and trust in automated decision-making processes, particularly since many AI models still operate as black boxes that are difficult to interpret and audit. This study investigates the potential of integrating blockchain technology to enable trustworthy and transparent AI decision-making and is conducted under the framework to systematically design, implement, and evaluate the proposed solution. The proposed framework records AI inference results and relevant metadata onto the blockchain through smart contracts to ensure data immutability and traceability. A prototype system is developed and evaluated using a mixed-method approach, combining qualitative analysis of transparency and auditability with quantitative measurements of system performance such as latency and overhead. The results demonstrate that blockchain integration significantly enhances auditability, data integrity, and user trust compared to conventional AI systems. However, several limitations are identified, including scalability issues, transaction costs, and increased latency caused by on-chain recording processes. Despite these challenges, the proposed approach shows strong potential to improve the accountability of AI systems in high-risk environments and contributes a practical framework along with empirical insights for organizations seeking to adopt transparent and reliable AI, while also opening opportunities for further development through architectural optimization and the adoption of layer-2 blockchain technologies.
Co-Authors Chua Toh Hua Dwi Cahyono Ikyboy Van Versie Rizki Indrawan Rohim Rohim