Clustering Text Documents Using High Utility Itemset Mining

The focus of the study in this paper is on clustering text documents using high utility itemset mining. The proposed algorithm groups two documents based on the number of high utility words which are common to both the documents. The utility of a word in a document reflects the frequency of occurrence of the word in the document. High utility words are the once having high frequency of occurrence. Utility of a word can be influenced by the section of the document to which it belongs. Each section is assigned weight depending on the importance of the section. Depending on the section to which words belong, although not frequent, they reflect the content of document better than some frequent word. Such words are given more weight so that they belong to the set of high utility words. Therefore, in this paper, the utility of a word is expressed as a function of its frequency of occurrence and the weight of the section it belongs to.