Research On User Perspective Mining Methods In The Network Environment

In the network environment, information and data have been rapidly expand. In all kinds of information, the user information plays a very important role, user view on social network public opinion can be used in the investigation of mining, the e-commerce site for the user can view of digging for the product design and promotion of businesses provide valuable reference for users to view must be based on large data mining. Information on the user's point of view is contained in the user's potential interest, also contains the user's emotional state. In this paper, based on the Chinese Weibo product reviews the study of mining method of user views, found that plays a very important role in the potential consumer groups, businessman can develop more targeted product marketing strategy accordingly, so as to really play a big role in data mining.This paper is to research mainly includes the following contents:(1) discussed the user view of text analysis, mainly including chapter level user view of classification and sentence level user view classification and classification of user view of vocabulary, etc.(2) studied the advantages and disadvantages of all kinds of classification algorithm, and through the instance of random forest algorithm and support vector machine algorithm in generalization ability, noise robustness and unbalanced classification on the similarities and differences, so as to choose the article will use the classification algorithm.(3) build based on syntactic weibo users view of dependency analysis model and weibo emotion analysis model based on text classification, and analysis and experiment.(4) System test, mainly including the construction of the test environment, Hadoop platform test and test results analysis, etc.This paper put forward innovative and text classification based on syntactic dependency relation with the combination of Chinese weibo users view analysis algorithm, the test results show that the algorithm accuracy and recall rate can reach nearly 90%, realize the user view analysis system has strong reliability and high flexibility, easy to extend, can achieve rapid screening of massive weibo data, it can be applied in all kinds of social networking and e-commerce sites.