Abstract
Ontology-driven conceptual models are typically regarded as a distinct type of conceptual models that leverages the semantics of foundational ontologies to provide a strong basis for modeling elements. Designed as tools for knowledge sharing, these models are expected to be extensively reused within their respective domains. However, studies indicate that understanding existing artefacts often poses challenges for users. We approached this problem from two perspectives: model abstraction and model explanation. First, we developed an algorithm for abstracting models grounded in Unified Foundational Ontology and specified in its associated modeling language, OntoUML. This algorithm generates a simplified version of the original model, retaining its essential information by utilizing the ontological semantics of the foundational ontology. The quality of these abstractions was evaluated using the FAIR catalog of OntoUML models and through several user studies. Second, we hypothesized that the process of model understanding resembles an information retrieval process. We assumed that when the latter is well-organized, it can improve model comprehension, enabling faster understanding with fewer errors. To support users in this process, we proposed several model explanation transformations aimed at enhancing the exploration experience. Finally, we examined the idea that specific conceptual model views can act as explanations for particular exploratory questions. We demonstrated how these views and questions can be systematically constructed. Our findings reveal that the pattern-based approach facilitates the creation of model views with fewer elements than the original model, while still adequately addressing the targeted question. The outcomes of our research are reflected in several scientific publications and have been implemented in the “ExpO” prototype system, which assists users in exploring OntoUML-based conceptual models.