Font Size: a A A

Application Of Data Mining Technology In Q&A Communities

Posted on:2019-04-18Degree:MasterType:Thesis
Country:ChinaCandidate:F X WangFull Text:PDF
GTID:2416330566499460Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
With the development of the Internet and the geometric expansion of knowledge and information,it is rapidly becoming the hottest way for people to exchange information through exchanging experiences,exchanging information and getting information from peers and experts through online Q&A communities.As a result,the online Q&A community platform has accumulated rich and extensive user insights and expert knowledge covering all aspects of business,culture and life and has therefore important Datamining value in decision-making and management fields which based on information and knowledge such as business,social research and government management.This paper chooses Zhihu which is the most influential online Q&A community in China as the research object and proposes a system and method of topic portrait based on the content creation and dissemination mechanism.In this paper,we use empirical research methods and select the topics of business and social studies as the representative of topic portrait application.Through compilation of the Web Crawler,we get 2533 questions,341505 user answers,382606 user information and 39266 topic information.Furthermore,aiming at the high dimension of question and topic data,this paper proposes a method of question-based opinion and motivation classification,a method of user topic tag customization;Aiming at the user portraits in the topic portraits,this paper proposes a user portrail method and model which includes user activity,user ability value and user interest degree;Aiming at the network data in the topic portrait,this paper proposes the mutual visualization and network analysis data mining method for the co-occurrence network of the problem parent topic;Aiming at the text data in the topic portrait,a data mining method based on topic mining and subject preference difference analysis for users' questions is proposed.In the end,this paper completes the empirical analysis by means of independently programming of Web Crawler,data preprocessing and data analysis,web visualization development,demonstrating the application value of topic portrait systems and methods in the field of business and social studies and the result of the reseatch provides a reference value for the managers or social researchers.
Keywords/Search Tags:Online Q&A community, data mining, topic portrait, statistical analysis, network analysis, text mining
PDF Full Text Request
Related items