Classifying sports news to improve search engine using web ontology language(OWL)

Abstract

The amount of information on the internet is growing rapidly and searching relevant content is center of concern nowadays. With the growing demand for relevant search results, it is important to introduce an approach which can transform implicit knowledge available on websites explicitly available to the users. Semantic Web is gaining popularity as it mainly focuses on learning the content meaning. Ontology is the basis of Semantic Web. In the following research, we present the retrieval of semantic information regarding Cricket and Football domain employing Ontology & SPARQL. In the initial step, two Ontology models are created for the Cricket and Football domain. News data of both targeted domains are collected from the Web in tabular format, and then preserved in the domain Ontology models. Web Ontology Language (OWL) is used for storing the two Ontology models. An algorithm is developed for creating triples. The SPARQL query language is applied on the two Ontology models for searching desired information and to make implicit information present on the website as explicit. Extracting information from a specific domain enhances the retrieval efficiency of the system. Evaluation of the two Ontology models is done using two performance parameters Precision & Mean Average Precision (MAP).