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

Ontology-Based Model Abstraction

“…On the other hand, this complexity poses a serious additional challenge in the comprehension and, consequent, quality assurance of these models. For example, [22] reports on an empirical study conducted with a large and professionally constructed conceptual model. 1 In that study, the authors managed to show that the model contained 879 occurrences of error-prone structures (anti-patterns), 52.56% of which really introduced representation errors according to the creators of the model.…”

“…As previously discussed, relators are fundamental in OD-CMs. For instance, by representing the cardinality constraints c 1 ..c n , d 1 ..d n , a 1 ..a n and b 1 ..b n in table I, we can eliminate instances of the so-called single-tuple/multiple-tuple cardinality ambiguity problem [12] in the model; by representing the the cardinality constraints e 1 ..e n and f 1 ..f n , we can eliminate occurrences of the repeatable instances anti-pattern as discussed in [22]. However, in order to compress our model, and following the previously discussed rationale, we abstract from this relator pattern by replacing each of its occurrences with a structure in which only the derived relations are shown.…”

“…Moreover, they cover a variety of domains ranging from Biodiversity (OpenBio) and Heart Electrophysiology (ECG) to Optical Networks (G.805) and Governamental Organizations (MPOG). For a discussion about the OntoUML model repository and these particular models, the reader should refer to [22]. Table II reports on the result of these tests.…”

“…11 For example, OntoUML is supported by an approach of model validation and anti-pattern detection by visual simulation [2], [22]. Two limitations of that approach are: the computational complexity of the simulation process -which relies on constraint satisfaction over a SAT solver; the cognitive tractability of the model instances generated by these simulations.…”

Considerations regarding Ontology Design Patterns

“…A thorough and evidence-supported understanding of these issues, e.g. what exactly "poor design" and "good design" are, appears to be necessary to advance on these fronts; investigations into ontological anti-patterns fall into this realm [22]. In particular, there seems to be a lack of user-centric evaluations addressing the ontology reuse issue, with only a few notable exceptions, e.g.…”