Title: Automatic relation extraction using naïve Bayes classifier for concept relational ontology development

Authors: G. Sureshkumar; G. Zayaraz

Addresses: Department of Computer Science, Pondicherry University, Karaikal-609605, India ' Department of Computer Science and Engineering, Pondicherry Engineering College, Puducherry-605014, India

Abstract: In this paper we proposed a methodology to learn concept relations from the unstructured text using dependency parsing pattern-based hand coded rules and to automatically construct domain ontology using extracted concept relations. The pattern mining was achieved using a recursive binary decision engine. We used syntactic and semantic probabilities obtained from the WordNet similarities as the concept relation features to train the naïve Bayes classifier designed to learn concept relations. The naïve Bayes classifier is bootstrapped by using an expectation maximisation procedure. The experiment conducted using benchmark datasets produced promising results. The effectiveness of the proposed method was proved by comparing the performance with similar well performing automatic ontology construction methods.

Keywords: relation extraction; concept extraction; ontology learning; ontology development; computer aided concept modelling; naive Bayes classifier; concept relational ontology; concept relations; unstructured text; dependency parsing; hand coded rules; pattern mining.

DOI: 10.1504/IJCAET.2015.072599

International Journal of Computer Aided Engineering and Technology, 2015 Vol.7 No.4, pp.421 - 435

Received: 25 Apr 2013
Accepted: 24 Jul 2013
Published online: 22 Oct 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article