Your browser does not support JavaScript!

Home    Σχεδιασμός και υλοποίηση συστήματος σημασιολογικής συσχέτισης σχημάτων  

Results - Details

Add to Basket
[Add to Basket]
Identifier 000356805
http://elocus.lib.uoc.gr
Title Σχεδιασμός και υλοποίηση συστήματος σημασιολογικής συσχέτισης σχημάτων
Alternative Title Design and implementation of a system for semantic schema matching
Author Καλαϊτζάκη, Μαρία Γεωργίου
Thesis advisor Πλεξουσάκης, Δημήτριος
Abstract The impact of schema matching in the domain of data integration is considered to be a fundamental importance, due to the extensive variety of its applications. Schema matching is the process of developing semantic matches between two or more schemas, and the detection of semantic correspondences between them. Despite the fact that the problem of schema matching is recurring and many solutions have been proposed, there is still the need of a unified system of a schema matching algorithm and a human-centric user interface. The core of this approach is described by the combination of the schemas’ lexical, semantic and syntactic information. The proposed solution utilizes the results of lexical matching, through the external dictionary WordNet, and the attributes of each concept, in order to produce the semantic similarity. Afterwards, the syntactic similarity is computed based on the relations of inheritance. Finally, these three factors decide the overall similarity between the concepts of the two schemas. The principal outcome of this work is an integrated system based on a hybrid semantic schema matching algorithm and a user-friendly web interface. The universality of this approach depends not only on the variety of the different types of schemas (relational schemas, ontologies and XML schemas) but also on the many factors that are considered in order to produce the similarity between two concepts. The experiments that have been carried out demonstrate the effectiveness and importance of the results.
Language Greek
Issue date 2009-06-23
Collection   School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses
  Type of Work--Post-graduate theses
Views 495

Digital Documents
No preview available

Download document
View document
Views : 9