Collins Obiorai Ibezim
Research Subjects
Collins Obiorai Ibezim has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Chibuzo Benjamin Onuike
Department of Mechanical Engineering, Federal University of Technology Owerri, Nigeria
1 shared publication Ginika Patricia Oguchi
1 shared publicationMichael Oluwaseyi Aderibigbe
1 shared publication- Olakekan Akinbosoye Okewale
Department of Computer Science, Ajayi Crowther University, Oyo, Nigeria
1 shared publication - Segun Kazeem Fatoki
Department of Computer Science, Ajayi Crowther University, Oyo, Nigeria
1 shared publication Temitope Dickson Olajide
1 shared publication
Latest Publications
- Open Access
- 9 December 2025
- Published by Mediterranean Publications and Research International in International Journal of Science Research and Technology
THE CONVERGENCE OF MACHINE LEARNING AND DECENTRALIZED APPLICATIONS: A PARADIGM SHIFT IN CYBERSECURITY PROTOCOL
The combination of machine learning (ML) with decentralized applications (dApps) signifies a major shift in the cybersecurity landscape, introducing both enhanced defensive capabilities and new categories of vulnerabilities. This paper presents an in-depth examination of this convergence, analyzing how ML techniques—including anomaly detection, behavioral analysis, and adaptive response models—can significantly strengthen the security posture of dApps. Using a systematic literature review and a design science research approach, the study demonstrates that ML enables a proactive and intelligent security framework for decentralized systems, achieving an estimated 20–35% improvement in threat detection accuracy as reported across reviewed implementations. The analysis also identifies a widened attack surface, particularly concerning data poisoning risks in federated learning, adversarial examples targeting on-chain inference models, and oracle manipulation vulnerabilities. In response, the paper proposes a multi-tiered security model incorporating privacy-preserving mechanisms, blockchain-based governance structures, and adversarial-resilient ML pipelines. While the abstract is concise and effectively captures the manuscript’s essence, it is strengthened by integrating specific evaluative insights that highlight the measurable benefits and risks of ML-enabled decentralized intelligence. The study underscores the need for security-centric design principles and offers actionable insights for developers, security professionals, and policymakers working to advance cybersecurity in intelligent, autonomous decentralized systems.
- 9 December 2025
- Published by Mediterranean Publications and Research International in International Journal of Science Research and Technology
Open AccessTHE CONVERGENCE OF MACHINE LEARNING AND DECENTRALIZED APPLICATIONS: A PARADIGM SHIFT IN CYBERSECURITY PROTOCOL
The combination of machine learning (ML) with decentralized applications (dApps) signifies a major shift in the cybersecurity landscape, introducing both enhanced defensive capabilities and new categories of vulnerabilities. This paper presents an in-depth examination of this convergence, analyzing how ML techniques—including anomaly detection, behavioral analysis, and adaptive response models—can significantly strengthen the security posture of dApps. Using a systematic literature review and a design science research approach, the study demonstrates that ML enables a proactive and intelligent security framework for decentralized systems, achieving an estimated 20–35% improvement in threat detection accuracy as reported across reviewed implementations. The analysis also identifies a widened attack surface, particularly concerning data poisoning risks in federated learning, adversarial examples targeting on-chain inference models, and oracle manipulation vulnerabilities. In response, the paper proposes a multi-tiered security model incorporating privacy-preserving mechanisms, blockchain-based governance structures, and adversarial-resilient ML pipelines. While the abstract is concise and effectively captures the manuscript’s essence, it is strengthened by integrating specific evaluative insights that highlight the measurable benefits and risks of ML-enabled decentralized intelligence. The study underscores the need for security-centric design principles and offers actionable insights for developers, security professionals, and policymakers working to advance cybersecurity in intelligent, autonomous decentralized systems.