Suspicious Model Detection to Improve Model-Based System Development
2026 (English)In: Communications in Computer and Information Science, Springer Nature , 2026, p. 27-38Conference paper, Published paper (Refereed)
Abstract [en]
Model-Driven Engineering (MDE) enhances software development by utilizing abstract system models. However, the existence of anomalous models can compromise system quality and reliability. This paper introduces an unsupervised machine learning approach that employs the Isolation Forest algorithm to detect anomalies in software models without requiring labeled data. The method achieves a 99% accuracy on Ecore models and 98% accuracy on UML models, demonstrating its effectiveness in identifying anomalous models and enhancing the robustness of model development.
Place, publisher, year, edition, pages
Springer Nature , 2026. p. 27-38
Series
Communications in Computer and Information Science, ISSN 1865-0929
Keywords [en]
Anomaly Detection, Isolation Forest, Machine Learning, Model-Driven Engineering, ModelSet, Forestry, Labeled data, Learning algorithms, Learning systems, Software design, Unsupervised learning, Abstract systems, Machine-learning, Model-based system development, System models, System quality, System reliability
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:mdh:diva-76929DOI: 10.1007/978-3-032-16808-5_3Scopus ID: 2-s2.0-105039278981ISBN: 9783032168078 (print)OAI: oai:DiVA.org:mdh-76929DiVA, id: diva2:2063121
Conference
31st International Conference on Information and Software Technologies, ICIST 2025, Kaunas, Lithuania, 16-17 October, 2025
2026-05-282026-05-282026-07-08Bibliographically approved