Ontology Learning Procedures Based on Web Mining Techniques

Abstract

Ontology establishes an approach for knowledge representation, sharing and reusing across various applications. It is also the backbone of the semantic web. Therefore, there is a requirement of rapid and effective ontology construction methodology for developing sophisticated knowledge-based and semantic web-based system. However, setting up ontology manually is a tough, time consuming, error-prone, costly, and inflexible task which requires the significant participation of a knowledge engineer or domain expert. An automatic or semi-automatic ontology construction methodology will be a better solution to this problem. Over the last decade, some attempts have been made by the researchers that provide assistance to ontology engineer to build ontologies. Still, there are several open issues in this field. The paper presents a brief overview of different tasks in relation to ontology learning method and reviews some of the ontology learning frameworks or methodologies based on web mining techniques. The comparative analysis of them in the tabular form based on the distinguishing ontology learning tasks has also been presented. Furthermore, we have discussed some of the recent problems and major drift related to the construction of ontology from semi-structured web pages.

Keywords: Ontology Learning, Semi-Structured Web Pages, Web Log Files, Web Mining, World Wide Web