Semi-Automated Ontology building through Natural Language Processing

Ontology is a backbone of semantic web which is used for domain knowledge representation. Ontology provides the platform for effective extraction of information. Usually, ontology is developed manually, but the manual ontology construction requires lots of efforts by domain experts. It is also time consuming and costly. Thus, an approach to build ontology in semi-automated manner has been proposed. The proposed approach extracts concept automatically from open directory Dmoz. The Stanford Parser is explored to parse natural language syntax and extract the parts of speech which are used to form the relationship among the concepts. The experimental result shows a fair degree of accuracy which may be improved in future with more sophisticated approach.