Research On Text Semantic Orientation Analysis For Areas Of Applied Based On The Web Information

Along with the computer and the rapid development of Internet technology, the network has become an indispensable access to information sources. Since Internet access Web2.0 era, more and more accustomed to the network as users express their views, ideas, attitudes of the platform, not just passively accept the site's information. Since a large number of users involved in the production of information, content and form of network information has become increasingly diverse, a lot of the contents of a personal point of view of the full network. These ideas for e-commerce, network and information security, network and other aspects of public opinion is very important significance. Orientation of the text is through the collection, organization and analysis of the position of these texts, ideas, opinions, attitudes, emotions and other subjective information to make the text tend to appraise the emotional judgments. This technology can be widely used in product reviews analysis, the attitude of weibo, social public opinion analysis, question and answer analysis and evaluation and other aspects.This paper comments on an Internet message based on the text for the study, the use of information collection, text processing, by means of data mining, computational linguistics and other fields of theory and method of application of the field study, carried out the modeling text orientation, Analysis and calculation of, explore the text orientation of the new technologies and new methods will be applied to specific areas, and achieved good results. This article includes the following aspects:First, this paper describes the research background of the text orientation and significance, research status, and described the tendency of network information flow of the text and the key technology.Second, the system introduces the knowledge of text classification, and analysis based on both statistical and semantic analysis of text orientation approach, and the tendency of the two main approaches to the text summarizes the advantages and disadvantages.Then on this basis was proposed based on pattern matching based on the extraction and text classification algorithms, and use two open corpus:corpus ChnSentiCorp Chinese emotional excavation and evaluation corpus COAE Chinese emotional evaluation of the algorithm, the algorithm to maintain a high performance The same time get a relatively good accuracy.Then, the proposed algorithm for technology and books on Dangdang comments, Sina Weibo Shakespeare event, Sina blog Libyan incident, the Foreign Ministry regular press conference Q & Record analysis of data, text orientation, dig out the comment and the length of The relationship between emotional tendencies, and users change over time of an event on the same change of attitude.Finally, the paper summarizes the research work proposed in the future direction of further research.