抽象的な

A Semantic Model for Concept Based Clustering

S.Saranya, S.Logeswari

Text mining is the data extraction from textual databases or documents. Each word in the document is a dimension giving a structure to the data and reducing the dimensions. In text mining techniques the basic measures like term frequency of a term is computed to work out the weight of the term in the document. Although with the statistical analysis, the original meaning of the term may not take the precise meaning of the term. The proposed system relies on concept based model. In this concept based approach, the concepts are extracted from the documents, and a semantic based weight is computed for effective indexing and clustering. It uses MeSH ontology for concept extraction and concept weight calculation based on the identity and synonymy relations. K-means algorithm is used for clustering the documents depending on the semantic similarity. Experiments are conducted and the results are analyzed.