The aim of this paper is to employ the principles of data mining and classify a new algorithm. In this method, we have proposed a new anti-spyware system (Spyware Detection), which is capable for classifying spyware files from legitimate files. Also we have evaluated the performance of our anti-spyware system with the existing anti- spywares in terms of overall accuracy (ACC) and false positive rate (FPR)... This paper describes the use of classi?cation trees and shows the methods of pruning them using the new Algorithm.The RIM algorithm which can be classified and compared with other algorithms like Naïve Bayes, HNB (Hidden Naïve Bayes) , Random Tree Algorithm The experimental results suggest that our method is better than the existing methods and shows the best result.