The analysis and optimization of KNN algorithm space-time efficiency for Chinese text categorization

Abstract

The performance of any algorithm for text classification are reflected in the of reliability classification results and classification algorithm is high efficient. We analyze the space-time efficiency of different stages based on the traditional KNN algorithm process for Chinese text classification and ensure the reliability of classification. And we optimize efficiency of the algorithm and the feasibility in the practical application from these aspects including feature extraction, feature weighting, similarity computing etc. © 2011 Springer-Verlag Berlin Heidelberg.