Documents Clustering Based on Optimized Compressibility Vector Space

To access and store large-scale electrical documents becomes possible due to the high performance of computer hardware and broadband accessible network. In order to handle these increasing number of documents properly, a efficient document representation model is as important as the classification algorithms. Several text representation methods, such as bag-of-words and N-gram models, have been widely used. Another representation approach named pattern representation scheme using data compression (PRDC) has been proposed lately. It does not only independently process data of linguistic text, but also processes multimedia data effectively. In this study, we will propose a method to improve PRDC approach and compare it with the two aforementioned methods. The performances will be compared in terms of clustering ability. Experiment results will show that the proposed method can provide better performance than that of the other two methods and also the PRDC.