This study proposes a design methodology for a self-evolving system that can proactively detect failures occurring in educational information systems and automatically implement corrective measures. Because educational information systems exhibit highly regular and cyclical structures—driven by academic schedules and recurrent user activity patterns. To this end, the study designs an anomaly-detection framework that combines Poisson-distribution–based frequency analysis with regular-expression–based pattern grouping. Detected anomalous events are forwarded to an automated deployment process that generates and applies corrective logic, which is then incorporated into active service instances without downtime through suggested deployment structure. This study presents a reliability-enhancement model that shifts failure management in educational information systems from reactive recovery to a proactive and automated paradigm. In experiments involving 79 stability test cases, the proposed system achieved 88.5% precision in anomaly detection and a 94.9% success rate in automated deployment, demonstrating its practical effectiveness in improving system stability. The approach is broadly applicable not only to educational platforms but also to a wide range of public and private information-system environments.
목차
Abstract 1. Introduction 1.1 Purpose of the study 1.2 Research Gap and Research Purpose 2. Related Work 2.1 Research Trends in Educational Information Systems and LMS 2.2 Research on System Failure Analysis and Anomaly Detection 3. System Design Methodology 3.1 Design of Log-Based Statistical Pattern Analysis (Detection Architecture) 3.2 Design of the Mitigation and Automated Validation Architecture 4. Functional Experiment and Analysis 4.1 Statistical Anomaly Detection Experiment 4.2 Automated Mitigation and Stability Verification 5. Conclusion and Discussion References
키워드
education information systemlearning management systemsfeedbackPoisson ModelingLog-based AnalysisAnomaly DetectionDesign PatternsAutomated Deployment; Online learning management systems
저자
Changbae Mun [ Professor, Department of Computer Science Engineering, Hanyang Cyber University, Korea ]
Corresponding Author