Multi-layered knowledge representation for Semantic searches

Abstract

Ontologies are the concepts that form the basis of a knowledge representation system which captures the semantic information of a domain. Construction of ontologies from textual information by extracting facts and patterns is becoming increasingly common. The objective of this research is to automate the process of constructing such a knowledge representation system using domain ontologies. In this study, universities are chosen as the domain to construct ontologies that can be used for interesting applications and better information retrieval. Wikipedia terms and categories are used as a dictionary for extracting information from university webpages. The resultant concepts of each university are analyzed and validated by predicting current research collaborators and competitors for a given domain expertise. It is hoped that the domain knowledge obtained from the result of this study can be used to predict potential research collaborators and competitors.