Concept Relevancy Improvement for Ontology Building through Term's Relations Extraction

Iram Shahzadi, Muhammad Shahbaz, Waqar Mahmood, and Ishtiaq Hussain


ontology, concept refinement, concept ranking, term frequency, ontology refinement, ontology generation


Ontology organizes and represents knowledge within a domain. It captures concepts and relations among concepts, which make the available information sharable and reusable. Manual creation of ontologies is a laborious activity therefore it is beneficial to have some automatic means to acquire ontologies. Domain specific ontologies may be generated automatically or semi automatically, if concepts and their relations could be extracted from given domain related text. In ontology generation process, relations between different concepts can play important role for the refinement of extracted concepts, but are generally ignored. In this paper we propose that while finding the relevancy of the concepts in ontology building process, we can consider relations between the extracted concepts and rank these concepts according to number of associated relations. Concepts, with greater number of associated relations are more relevant and more valuable. Hence the generated ontology, based on relation counting, is more precise.

Important Links:

Go Back