E-Locus - Institutional Repository of the University of Crete

Home Search

Results - Details

Search command : Author="Φλουρής" And Author="Γεώργιος"

Current Record: 15 of 22

[Add to Basket]

Identifier

000386820

Title

A diagnosis and repair framework for DL-LiteA knowledge bases

Alternative Title

Πλαίσιο εντοπισμού και επιδιόθρωσης λαθών σε γνωσιακές βάσεις DL-LiteA

Author

Χόρτης, Μιχάλης

Reviewer

Φλουρής, Γεώργιος
Πλεξουσάκης, Δήμήτριος

Thesis advisor

Χριστοφίδης, Βασίλης

Abstract

Several logical formalisms have been proposed in the literature for expressing structural and semantic integrity constraints of Linked Open Data (LOD). Still, the data quality of the datasets published in the LOD Cloud needs to be improved, as published linked data often violate such constraints. This lack of consistency may jeopardise the value of applications consuming linked data in an automatic way.A major challenge in this respect, is to provide to the curators of linked data knowledge bases (KBs), the tools that will help them in detecting the violations of integrity constraints and in resolving them, in order to render the knowledge base valid and improve its data quality. In this work, we propose a novel, fully automatic framework for detecting violations of integrity constraints (diagnosis) in KBs, by executing the appropriate queries over the data, as well as for resolving those violations (repair ), by removing invalid data from the KB. Our approach takes into consideration both explicit and inferred ontology knowledge, by relying on the ontology language DL-LiteA for the expression of several useful types of logical constraints and for the detection of data that are inconsistent with those constraints, while maintaining good computational properties. The framework that is proposed in this work is modular, allowing each component to be implemented in a manner independent to the other components. This way, we are able to implement our framework with using off-the-shelf, state-of-theart tools for several features, such as reasoning, query execution, etc. We have implemented and evaluated our framework, showing that it is scalable for large datasets and numbers of invalidities, which are exhibited in reality by reference linked datasets, such as DBpedia. The evaluation also shows that our framework can be used over already deployed knowledge bases, without any further reconfiguration.

Language

English, Greek

Subject

DL-LiteaA Constraints

Diagonis

Integrity constraints

Repairing

Διάγνωση

Επιδιόρθωση

Περιορισμοί DL-LiteA

Περιορισμοί ακεραιότητας

Issue date