Research Of Internet Fine Grit Opinion Mining Based On The Dependence Syntactic Parsing

With the rapid growth of Internet users and a growing range of network information platform, forum, blog, personal home pages and BBS have become concerned about the people in the network of information and comments on the importance of media. Network on the critics, book reviews and comments on certain products and services is also rapidly increasing. Thus, when companies or consumers need to obtain valuable information on the comments, they mass is difficult to comment in the network quickly and comprehensively to find the information you want. At this point opinion mining had a huge role, it automatically on the network for a large number of unstructured text mining to quickly get valuable ideas and attitudes. Many foreign experts and scholars on the opinion mining to do some useful research, and related systems have been put into use. But because the Chinese language processing in the machine there are some characteristics and difficulties of its own, so the domestic point of view of research is still in the mining exploration and development stage. In recent years, many of the experts and scholars have made some progress and breakthroughs.The traditional view of subjective opinion mining can type in text analysis and processing point of view obtained after the holder appraise attitude objects that comment on was the attitude of the praise, derogatory or neutral. But this technology can not meet people's growing demand for in-depth view mining. Thus, in recent years, research began to tap the ideas tend to fine-grained mining, that is not only able to get the public's attitude to appraise better identify which product is which features of the view.Based on the point of view applied to the excavation process and the main technical approach, introduced ideas of web mining and pre-crawling technology, Chinese word segmentation based on feature-based view emotional and mining methods. Subjective point of view-based text has four elements:theme, point of view holder, and emotional state, this paper introduces the theme, the holder of the extract and view the classification based on emotion. Then dig for fine-grained details of the syntactic point of view analysis, and proposed to achieve fine-grained view mining process, some have made good use of the effect of analytical tools, such as:ICTCLAS tools and DeParser Chinese word parser, the other prepared for the dependency of the statistical analysis procedures. Then the word of the characteristics and polarity matching algorithms are introduced and improved. Finally, through experiments and data analysis point of view of fine-grained mining methods to achieve the necessary verification.Because mining is a fine-grained view at the forefront of research and the limited capacity of individual researchers, this article there are still some shortcomings, the need for further research to explore., such as the problem of precision of Chinese word segmentation and syntactic parsing, connection of circuit, and so on.