A framework for information integration using ontological foundations

Asgharzadehsekhavat, Yoones (2014) A framework for information integration using ontological foundations. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (2MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

With the increasing amount of data, ability to integrate information has always been a competitive advantage in information management. Semantic heterogeneity reconciliation is an important challenge of many information interoperability applications such as data exchange and data integration. In spite of a large amount of research in this area, the lack of theoretical foundations behind semantic heterogeneity reconciliation techniques has resulted in many ad-hoc approaches. In this thesis, I address this issue by providing ontological foundations for semantic heterogeneity reconciliation in information integration. In particular, I investigate fundamental semantic relations between properties from an ontological point of view and show how one of the basic and natural relations between properties – inferring implicit properties from existing properties – can be used to enhance information integration. These ontological foundations have been exploited in four aspects of information integration. First, I propose novel algorithms for semantic enrichment of schema mappings. Second, using correspondences between similar properties at different levels of abstraction, I propose a configurable data integration system, in which query rewriting techniques allows the tradeoff between accuracy and completeness in query answering. Third, to keep the semantics in data exchange, I propose an entity preserving data exchange approach that reflects source entities in the target independent of classification of entities. Finally, to improve the efficiency of the data exchange approach proposed in this thesis, I propose an extended model of the column-store model called sliced column store. Working prototypes of the techniques proposed in this thesis are implemented to show the feasibility of realizing these techniques. Experiments that have been performed using various datasets show the techniques proposed in this thesis outperform many existing techniques in terms of ability to handle semantic heterogeneities and performance of information exchange.

Actions (login required)

View Item