Abstract
dc:descriptionAs knowledge and data become more prevalent with the aim of achieving interoperability between distributed and decentralized systems, the problem of combining knowledge of individual reasoning chunks, and/or decomposing large ontologies becomes also important. This thesis contributes to:(a) Combining ontologies in expressive fragments of Description Logics in inherently distributed peer-to-peer settings with autonomous peers(b) Partitioning ontologies for efficient distributed reasoning(c) Providing a comprehensive toolset for distributed reasoning with distributed and/or large ontologies via modularization and combining local reasoning chunks in a peer-to-peer fashion.Specifically:(a) The E-SHIQ Representation framework for combining ontologies:Combining ontologies in expressive fragments of Description Logics in inherently distributed peer-to-peer settings with autonomous peers, is still a challenge in the general case. Although several modular ontology representation frameworks have been proposed for combining Description Logics knowledge bases, each of them has its own strengths and limitations. In this thesis we consider networks of peers, where each peer holds its own ontology within the SHIQ fragment of Description Logics, and subjective beliefs on how its knowledge can be coupled with the knowledge of others. To allow peers to reason jointly with their coupled knowledge, while preserving their autonomy on evolving their knowledge, data and subjective beliefs, we propose the E-SHIQ representation framework. E-SHIQ (i) provides constructors associated to link relations and correspondences between concepts of different ontologies, retaining the subjectiveness of specifications, (ii) offers expressiveness for combining multiple units that can be interpreted over overlapping and disjoint domains, (iii) supports distributed reasoning by combining local reasoning chunks in a peer-to-peer fashion, and (iv) inherently supports subsumption propagation between units. Each peer with a specific unit holds a part of a distributed tableau, which corresponds to a distributed model. The E-SHIQ distributed reasoner has been implemented as an extension of the Pellet reasoner. The thesis presents experimental results with the E-SHIQ distributed reasoner, on different distributed settings of particular interest. The results show that the reasoner achieves better performance than its centralized counterpart, although further optimizations may be applied.(b) Ontology modularization for efficient distributed reasoning:Ontology modularization methods aim either to extract modules from ontologies, or partition ontologies into sets of modules. Each module covers a specific part of the whole, and it should ``make sense'' as well as preserve additional properties. A variety of ontology modularization methods have been proposed to either extract modules from an ontology or partition an ontology into a set of modules, aiming to satisfy certain requirements depending on the modularization goals: Specifically, ontology modularization may aim at reusability of knowledge, reduction of complexity, efficient reasoning, and tooling support, e.g. for efficient ontology maintenance and evolution. This thesis proposes a generic framework for the modularization of SHIQ ontologies, towards the construction of distributed E-SHIQ knowledge bases. The aim is to compute decompositions for correct, complete and efficient distributed reasoning. The proposed framework combines graph-based modularization techniques with locality-based rules using a generic constraint problem solving framework. Experimental results show that the modularization task satisfies specific desiderata that contribute to increasing the performance of the distributed E-SHIQ reasoner (also in comparison to other well known reasoners and optimization methods).
Degree
thesis:*- Grantor dc:publisher
- University of the Aegean
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Santipantakis, Giorgos
- Σαντιπαντάκης, Γεώργιος
Subjects
dc:subject × 9Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/34646