Authors : R. Thilagavathy and R. Sabitha
Abstract: In text mining most of the methods are based on the concept of term (i.e., a word or a phrase) analysis. Statistical analysis usually identifies the important terms by means of their frequency within a document. However, more than one term may contain the identical occurrences within the document, however a specific term plays major role towards sentence semantics comparing to the remaining term. Therefore, the fundamental web document clustering method should specify term which identifies meaning of the text. In this case, the semantic-based method identifies expressions that represent the sentence semantics which are very helpful in determining the document’s subject. This mining model analyses words or expressions on the individual sentences, documents and core level. The semantic-based model dramatically distinguishes among insignificant terms against the meaning of the sentences and terms which are more close to the sentence semantics. The proposed method can effectively find important similar concepts among documents with respect to the meaning of their sentences. The interrelations among the documents are estimated by a similarity measure which is based on concepts. By using the semantic organization of sentences in the web documents, a considerable improvement in the quality of web document clustering is achieved.
R. Thilagavathy and R. Sabitha, 2016. Enhancing the Quality of Search Results by a Novel Semantic Model. Asian Journal of Information Technology, 15: 996-1004.