How do I find reusable models?

Models play a major role in model-based development and serve as the main artifacts that stakeholders aim to achieve. As it is difficult to develop good-quality models, repositories of models start emerging for reuse purposes. Yet, these repositories face several challenges, such as model representation, scalability, heterogeneity, and how to search for models. In this paper, we aim to address the challenge of querying model repositories by proposing a generic search framework that looks for models that match the intention of the user. The framework is based on a greedy search approach using a similarity function that considers type similarity, structure similarity, and label similarity. We evaluate the framework’s efficiency on different model types: UML class diagrams, Human Know-How, and ME maps. We further compare it with existing alternatives. The evaluation indicates that the framework achieved high performance within a bounded time, and the framework can be adapted to different modeling languages for searching related, reusable models.