Text Clustering And Its Application In Text Orientation Analysis

With the rapid development of the Internet,people increasingly rely on the Internet to obtain information,at the same time people can speak freely in the network,to express their feelings,published on things,even the evaluation of state affairs,legal opinion.It can be seen today,the network has become one of the most popular voice.Therefore,the government should make full use of the network,to understand the needs and opinions of ordinary people,timely solutions have appeared or possible problems,ease the negative emotions of netizens,harmonious society can truly achieve.Now,the government departments,as well as government agencies,and even some enterprises are through the sentiment orientation analysis system to get the attitude and perception of Internet users to treat a thing.At present,some of the emotional tendency analysis system uses different algorithms as the main algorithm of data analysis,supplemented by some text processing tools to achieve.Because of the difference of the algorithm,the accuracy of the analysis will be different.Based on the network text orientation analysis,presents a bounded,semi bounded non Gauss data has good fitting algorithm,is applied to the network text sentiment analysis,so as to realize a network system for text sentiment classification.1.Study the variational learning algorithm of infinite inverse Dirichlet mixture model(Infinite inverted mixture model,InIDMM).In order to solve the problem that the Gauss mixture model can not accurately fit the semi bounded and non-data,a InIDMM model is proposed for the accurate modeling of the semi-bounded data.The proposed algorithm can estimate both the model parameters and the number of mixed components simultaneously.Finally,the experimental results of synthetic data sets show the effectiveness of the proposed algorithm.2.Research and implementation of text orientation analysis system.In order to solve the problem of the difficulty in the analysis of the large text data in the web,a text orientation analysis system is proposed,which is based on the variational learning algorithm of the infinite inverse Dirichlet mixture model.The network text data acquisition,and through the web page denoising and a series of text preprocessing operation,finally this paper proposes an algorithm for text orientation analysis,so as to realize the overall process of the system.