The Implementation And Application Of The Emotional Tendency Text Classification Methods In Usefulness Reviews

Text classification as a large number of text data processing and organization of key technologies that can solve a greater extent of information clutter phenomenon of user-friendly and accurate information needed to locate and diversion information. In this paper, text classification and its related technologies were studied. By raising a reasonable classification, accuracy and stability, this paper proposes an effective solution or improvement of methods and techniques. The same time, the text classification technology, a new research direction - Text emotional tendencies classification, text classification is an important application areas - e-commerce sites on the usefulness of customer comments, were studied. This study content and improve our work include the following:1. research and summarized the characteristics of text classification and process emotions, analysis and comparison of several mainstream text classification emotional tendency to infer a machine learning based on the characteristics of text classification, and select such a method as a key research paper object.2. designed a machine learning-based text classification emotional tendencies. Described in detail the implementation of the method level and key modules, and at various levels and modules are given a specific design and implementation steps.3. presents a tendency to build a text classification emotional standardized sample set of common methods, detailed standardized sample set in the classification of the importance of implementation, the sample set of design principles, the basic structure and the specific build process.4. using the above method of constructing the sample collection and processing of e-commerce sites to collect the comments data, design experiments based on machine learning and semantic-based classification method associated with a specific implementation, and the two experimental results were analyzed and compared , then select the classification based on machine learning approach to the next comment on the usefulness of analysis.5. summing up the feelings of the general tendency of the text classification based on the principle of introducing a pair of the new text classification of emotional tendencies factor - the main, objectivity, and e-commerce sites on the customer comments on the usefulness as a research object, given the classification factor and comment on the usefulness of correlation between the mathematical model, design experiments to achieve the model and finally using statistical software to run the results are analyzed.