Context ontology driven relevant search using data mining techniques (CODT)

The size of the publicly indexable World Wide Web (WWW) has probably surpassed 14.3 billion documents and as yet growth shows no sign of leveling off. Search engines encounter the problem of ambiguity in words; therefore, search engines use ontology to find pages with words that are syntactically different but semantically similar. The knowledge provided by ontology is extremely useful in defining the structure and scope for mining web content. Context-ontology is a shared vocabulary to share context information in a pervasive computing domain and include machine-interpretable definitions of basic concepts in the domain and relations among them. This paper proposes an architecture for relevant searching of web documents using data mining techniques such as clustering and association rules. These techniques together with context and ontology extract potentially useful documents from the database. Also, an algorithm has been devised which shows the working in sequence of steps. Finally, the results are compared with the prevailing approaches and with the help of an example it has been seen that CODT is better in context of relevancy.