An approach towards Semantic Web Document Clustering using Background Knowledge

Extensive use of Information Technology in diverse applications has lead to massive online database. This database is often retrieved for various purposes. In order to smooth the progress of effective browsing and efficient searching for a user, one has to arrange the documents in a systematic manner. Clustering text documents is a vital step in organizing, management and indexing a huge text data on Web. Traditional approaches toil taking place in the keyword identical method, the routine destroys then the handler didn’t change to the suitable upshot. Therefore, it grows into essential to signify the article semantically besides formerly remain managed. The background knowledge: Wordnet can be used for this purpose. In this paper Semantic weight is used to represent the document semantically and then a nature inspired meta heuristic algorithm : Cuckoo Search is implemented for automatic web search results clustering. Wordnet concepts are been identified from the text by considering the ontology of the text word. A novel method to find the semantic similarity is presented to represent the document as features.. These enhanced features are fed to the clustering algorithm. Cuckoo Search algorithm solves the problem of automatically defining number of clusters. Divide n conquer technique is used to avoid algorithm converging too quickly. The algorithm is tested on AMBIENT Dataset and shown good results.