PPC:A novel approach of Chinese text mining based on Projection pursuit model

Abstract

A new fuzzy clustering algorithm (PPC) based on project pursuit model is presented in the paper. The main defect of the traditional clustering algorithm is to reduce dimension, while PPC don't need reduce dimension. Firstly, the text vector is normalized; Secondly, the project index function is constructed; Thirdly, the project function is optimized; Finally, the clustering result is acquired according to classification threshold. PPC algorithm improves the efficiency and precision of clustering.