Ontological anti-patterns: empirically uncovered error-prone structures in ontology-driven conceptual models

The construction of large-scale reference conceptual models is a complex engineering activity. To develop high-quality models, a modeler must have the support of expressive engineering tools such as theoretically well-founded modeling languages and methodologies, patterns and anti-patterns and automated supporting environments. This paper proposes a set of Ontological Anti-Patterns for Ontology-Driven Conceptual Modeling. These anti-patterns capture error-prone modeling decisions that can result in the creation of models that fail to exclude unintended model instances (representing unintended state of affairs) or forbid intended ones (representing intended states of affairs). The anti-patterns presented here have been empirically elicited through an approach of conceptual models validation via visual simulation. The paper also presents a series of refactoring plans for rectifying the models in which these anti-patterns occur. In addition, we present here a computational tool that is able to: automatically identify these anti-patterns in user's models, guide users in assessing their consequences, and generate corrections to these models by the automatic inclusion of OCL constraints implementing the proposed refactoring plans. Finally, the paper also presents an empirical study for assessing the harmfulness of each of the uncovered anti-patterns (i.e., the likelihood that its occurrence in a model entails unintended consequences) as well as the effectiveness of the proposed refactoring plans.