Review of feature selection methods for text classification

With the immense growth of online information due to Internet, text-categorization has become a very significant technology to classify a large number of documents to make them useful. The concept of text classification is to categorize textual data into one or more predefined categories 1-3. It is not possible to manage manually these big datasets. As a result, we witness the development of various machine learning and data mining methods. These methods are being used for automatic recognition of the patterns in large data sets for gaining useful information.