Evidence based similarity for document categorization

The failure of effective information management over the web does not influence only on by the speed of information retrieval; but also influenced by individual and organizational activities. Thus, the search engine of information retrieval plays a pivotal role in retrieving results that are relevant to users' query. Categorization is an optimist alternative to improve the accuracy and speed of information retrieval. This paper proposes a method using evidential reasoning based on similarity for text categorization. Term and frequency of term has influenced the process of deciding the document category. In order to define a new category for new document, our proposed method has taking degree similarities into consideration.