The Design and the Implementation of the Public Opinion Analysis System Based on Topic Discovery
Huang Mei-xuan · Journal of Beijing Union University · 2012
A set of Chinese word segmentation,vector space model(VSM) and feature extraction is developed to reduce the dimensions and text clustering and other modules of public opinion information analysis system and to realize public opinion topic discovery.In public opinion analysis system implementation process,a K-means algorithm is proposed.Experimental results show that the algorithm enhances the recall ratio and the precision ratio and gains a high and stable accuracy rate.