Main Article Content
Abstract
The World Wide Web grows at each fraction with number of documents. Such growth introduces challenge in clustering the documents. There are number of clustering algorithms has been discussed earlier but suffers to achieve clustering efficiency. To overcome the deficiency, the proposed algorithm introducedan efficient clustering algorithm which consider the relevancy of documents to be measured with internal and external documents. The method first computes the informatic similarity measure with all clusters and selects a higher one. In the second stage, the method compute the internal informative similarity and external informative similarity to compute the Informative weight. Based on computed informative weight the method assigns the class label for the web document. This algorithm uses 5,00,000 web documents for evaluation and 70 percent as training set and 30 percent as testing set. The method produces higher classification accuracy with less time complexity.