Ontology Materialization by Abstraction Refinement in Horn SHOIF

Abstract

Abstraction refinement is a recently introduced technique using which reasoning over large ABoxes is reduced to reasoning over small abstract ABoxes. Although the approach is sound for any classical Description Logic such as SROIQ, it is complete only for Horn ALCHOI. In this paper, we propose an extension of this method that is now complete for Horn SHOIF and also handles role- and equality-materialization. To show completeness, we use a tailored set of materialization rules that loosely decouple the ABox from the TBox. An empirical evaluation demonstrates that, despite the new features, the abstractions are still significantly smaller than the original ontologies and the materialization can be computed efficiently.

A Set of Conceptual Guidelines for Reverse Engineering Using Ontology-Based Materialized Views

“…Related work. As previously emphasized, only [9,10,11,13,14] tackled ontologies' materialization. Thus, we concentrate ourselves on these closely related works to ours and give a precise insight in the following.…”

“…Thus, we concentrate ourselves on these closely related works to ours and give a precise insight in the following. Globally, in [9,10,11], reasoning over "large ABoxes" by "abstraction refinement" was the focus. More precisely, arguing that "updates of ontology definitions are equivalent to the updates and new definitions of rules, whereas existing maintenance techniques only address the update of ground facts", these works "present a technique for the incremental maintenance of materialized ontologies".…”

“…Mainly, the above works are based on logic databases, using "Description Logic Programs (DLP)" as equivalent to the "OWL's Abstract Syntax" [9,10,11].…”

“…Indeed, ontologies allow the specification of high-level formal conceptualizations [3]. However, most of the proposed approaches, e.g., [2,3,4,5,6,7,8,9,10,11,12], are not domain independent. Therefore, data reuse is not easy to achieve or not achieved at all because of the common trend to approach the problem of reverse engineering with object-oriented techniques.…”

“…Indeed, according to [3], reverse engineering consists of producing a semantic model based on the code in which the data model or the knowledge model (e.g., an ontology) is implemented. Moreover, to the best of our knowledge, only [9,10,11,13,14] have dealt with the problem of ontologies' materialization. However, rather than the use of materialized views with ontologies for reverse reengineering, the authors only discussed ontologies' storage, without conceptualization of materialized views in the design process, and it is not clear how ontologies' models are linked to the business-data.…”