Research On The Tendeney Of Book Reviews Based On Web Text

With the popularity of Internet Web2.0, the Internet has gradually become the expression of emotions, opinions and share experience of new public space, more and more people are accustomed to on the network platform leaving their own views and opinions, which the Internet has More people are considered "Lecture."As part of this online bookstore e-commerce platform is the most important way people buy books, but also share the reading experience with its people, through books and on the evaluation of the "free environment", these divergent mass text book reviews have become books Public Opinion resource publishing, distribution and reading concern and research, a research issue in the field of Computer Applications, Library and Information Science, mass media and other interdisciplinary become big data analytics, data mining data support, has important research value. Users purchase goods online reviews of books will be decisions on when to buy their own goods. The contents of these books represent the feelings of the evaluation corpus generated in the book reviewer purchase or use of the process, many of them have a distinct theme and emotional tendencies, through the analysis of these data, we can find books from buyers from the use of Book angle, the value of the book to give an objective evaluation of the.Massive real text book reviews corpus for analyzing proposed a text from a multi-level analysis Solutions propensity, tendency of people to analyze the language of books evaluating large amounts of data, from which to extract the "Content", "quality", " logistics " three different themes evaluation corpus propensity to determine the book " Content " and " quality " value indicators under the three themes of " logistics. "Research method of this paper is, first, get a lot of book reviews reptile tools original corpus "online bookstore", and then, using keyword-based clustering method to construct the library evaluation Thesaurus. The book is designed according to different themes corpus reviews focusing algorithm, and finally, for the three themes comments by single sentence level, user-level to books and Sequential analysis algorithm designed to give a rule-based and statistical analysis algorithms tendentious book reviews on different themes books propensity analysis.This paper designs and implements a "topic-based classification of book reviews tendentious analysis system," the text of the algorithm is verified, draw better conclusions, the system also provides for future commercial applications feasible ideas.