Paper Title
Relevance Feature Discovery In Text Documents

Abstract
Text Mining is the process of deriving high quality information from text documents. High quality information is typically derived through the devising of patterns. The foremost hurdle faced in this regard is how to effectively use large scale terms and data patterns and resolve the issues of polysemy and synonymy. The conventional text mining approaches do not overcome these limitations. Hence this innovative model for relevance feature discovery makes use of large scale patterns and a method to find and classify low-level features based on both their appearances in the higher-level patterns and their specificity. The specificity function then categorizes the documents into relevant and irrelevant documents. The use of irrelevant feedback improves the efficiency of removing the noises from the document hence optimizing the text mining for the users. Keywords— Text Mining, Text Feature Extraction, Data Mining,Pattern Matching.