CYBER RISK ANALYSIS AND THREAT ASSESSMENTS FROM GOVERNMENT OPEN DATASETS ON EDUCATION USING HYBRID DEEP LEARNING MODEL: A LITERATURE REVIEW

Abstract

The institutionalization of open data policies and the digitization of educational systems have jointly produced large, publicly accessible data ecosystems that were not originally designed with cybersecurity analytics in mind. While government open datasets on education are widely used for transparency, planning, and research, their unintended role as indicators of cyber risk has received limited scholarly attention. This literature review examines how contemporary machine learning and deep learning approaches particularly hybrid deep learning models have been employed for cyber risk analysis and threat assessment, and evaluates their relevance to education-sector government open data. Drawing on peer-reviewed studies across cybersecurity, artificial intelligence, and open data governance, the review synthesizes methodological trends, model architectures, and empirical findings. The analysis demonstrates that hybrid deep learning frameworks integrating convolutional, recurrent, and attention-based components consistently outperform single-model approaches in cybersecurity contexts. However, their application to education-related open datasets remains sparse, indirect, and methodologically underdeveloped. The review identifies key technical, data-related, and governance challenges and argues for a shift toward explainable, adaptive, and policy-aligned cyber risk analytics tailored to education-sector open data environments.