Date Added: Mar 2012
Feature clustering is the way to reduce the dimensionality of features presents in the text documents and it is highly important for text categorization problems. The performance of the text classification is degraded when the dimensionality of input text is huge .Feature clustering is a powerful alternative to feature reduction approaches. The first task is to calculate the word patterns for each feature present in the text document. The second task is to calculate the membership function by the mean and deviation of the word patterns .The third one is to generate the clusters based on the membership function. Evaluation results for these tasks show that the proposed methodology obtains reliable performance for text classification tasks.