의미적 정보검색을 위한 온톨로지 자동 구축 및 효율적 질의 처리 방법

Abstract

As the study on the semantic web has been progressed, the applicable range of the ontology is getting wide. So, lots of ontologies were built and are exiting on the web nowadays. As many ontologies are made, several problems occured on the ontology side as well as on the semantic web. Ontology plays a primary role in the semantic information retrieval and several methodologies are needed to process the ontology efficiently. In this thesis, the automatic ontology building method, efficient ontology parsing method and triple-based ontology query processing method were studied. Firstly, the automatic ontology building method is designed based on the WordNet. The concepts collection, class creation and relation creation between classes among the ontology development steps are processed automatically. Secondly, the ontology parser is newly designed for the efficient ontology parsing. As a result, the parsing time was improved. The syntax parsing and semantic parsing about ontology are processed, and then two types of triple set are created throughout the parser. Finally, created triple sets are used as the input data of the ontology query processing method suggested. Therefore, the query processing method supports the semantic information retrieval about the ontology and user can write the queries easily in the query interface using the controlled query process module. Although the active study on the semantic web brings about the implementation of the ontology techniques, several methods for processing the ontologies efficiently are still demand. It should be studied about how to get the rich knowledge data from the domain experts or the others for the automatic ontology building and apply the query processing method for the ontologies to the information retrieval system.