Automatsko pronalaženje podudarnosti u ontologijama primjenom paralelne kompozicije osnovnih metoda

doctoral thesis

• Title (english): Automatic ontology matching by parallel composition of basic methods
• Author: Marko Gulić
• Mentor: Boris Vrdoljak (mentor)
• Committee member: Boris Vrdoljak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Applied Computing); Zagreb
• Defense date and country: 2014, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing; Process Computing
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Sve je veci broj heterogenih izvora podataka koji opisuju istu domenu interesa, a potrebno ih je povezati. Ontologija obogacuje znanje o izvorima podataka opisujuci entitete i njihove relacije unutar domene interesa i stoga njeno korištenje olakšava integraciju heterogenih izvora podataka. Kako bi integracija bila što uspješnija, potrebno je napraviti sustav za pronalaženje podudarnosti izmedu entiteta razlicitih ontologija. Disertacija se bavi automatizacijom pronalaženja podudarnosti izmedu ontologija. Fokus je na izradi metode za automatsko odredivanje težinskih faktora u težinskoj agregaciji paralelne kompozicije osnovnih metoda za pronalaženje podudarnosti. Takoder, izradena je metoda za automatsko odredivanje završnog poravnanja odgovarajucih parova entiteta usporedivanih ontologija. Implementiran je i prototip programskog sustava za pronalaženje podudarnosti izmedu ontologija u svrhu verifikacije izradenih metoda. Evaluacija prototipa programskog sustava izvršena je usporedbom prototipa programskog sustava s aktualnim programskim sustavima za pronalaženje podudarnosti izmedu ontologija. Rezultati evaluacije pokazuju da prototip programskog sustava uspješno pronalazi podudarnosti izmedu entiteta razlicitih ontologija.

Abstract (english)

The number of heterogeneous data sources that describe the same domain of interest is increasing and these sources need to be related. Ontology enriches the knowledge about data sources by describing entities and relationships between them within the domain of interest. Therefore, the use of ontologies facilitates the integration of heterogeneous data sources. In order to accomplish a successful integration, it is necessary to make an ontology matching system that will find similar entities among two ontologies. The dissertation deals with the automation of ontology matching. The focus is on the development of the method for automatic assignment of weighted factors in the weighted aggregation of parallel composition of basic methods. A method for automatically determining the final alignment of similar entities of compared ontologies is also made. The prototype of the software system for ontology matching is developed in order to verify proposed methods. The evaluation of the prototype was performed by comparing the prototype of the software system with current software systems for ontology matching. The evaluation results show that the prototype of the software system successfully finds correspondences between entities of different ontologies.

• Keywords: ontologija; integracija ontologija; pronalaženje podudarnosti izmedu ontologija; sustav za pronalaženje podudarnosti izmedu ontologija; osnovna metoda za pronalaženje podudarnosti; težinska agregacija; završno poravnanje
• Keywords (english): ontology; ontology integration; ontology matching; ontology matching system; basic matcher; weighted aggregation; final alignment
• Language: croatian
• URN:NBN: urn:nbn:hr:168:138133
• Study programme: Title: Electrical Engineering and Computing; Study programme type: university; Study level: postgraduate; Academic / professional title: Doktor znanosti elektrotehnike i računarstva (Doktor znanosti elektrotehnike i računarstva)
• Type of resource: Text
• Extent: 146 str. ; 30 cm.
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-04-17 15:49:38