Knowledge Network Based On Text Topic In Social Q&A Sites

User-generated content in social Q&A sites,as a knowledge corpus of user thoughts and needs,has been paid attention to by researchers in recent years,and social Q&A sites represented by Quora and Stack Overflow have become an effective way to bridge the user's real-time search for tacit knowledge and practical information.In social Q&A sites,the topic is a collection of domain-specific knowledge.If it is regarded as a unit of knowledge,the user's interaction behavior,information behavior and semantic association of user-generated content will connect different knowledge,to form different knowledge networks with rich information.Research on knowledge networks can help researchers and users better understand the process of dissemination and diffusion of knowledge in social Q&A sites to enable people to share,use,represent and extract knowledge more effectively.At present,researchers often explore the semantics,relevance and influencing factors of knowledge from the content characteristics of knowledge networks,or explore the static attributes and dynamics of networks from the topological characteristics of knowledge networks.There are few studies to combine the two aspects,or combine multidisciplinary theories and ideas to conduct research.Therefore,this paper proposes a knowledge network model based on text topic,in order to combine the topological characteristics and non-topological characteristics of knowledge networks to deeply explore the knowledge corpus in social Q&A sites.This research is carried out in order of logic level.Firstly,based on the research of social Q&A sites,it analyzes the knowledge exchange and organization in social Q&A sites,and focuses on the platform information organization logic to provide a framework for the extraction,integration and analysis of knowledge units.Based on the characteristics of LDA model and knowledge network,a knowledge network model based on text topic is proposed.Based on this,the static topological characteristics and dynamic properties of knowledge network are analyzed.At the same time,it conducts research on the sub-community of Stack Exchange,a world-renowned social Q&A site,and demonstrates the feasibility and effectiveness of building a knowledge network model through empirical verification.Research shows that(1)social Q&A sites have a complete "topic-question-answer"structure,users can comment on questions and answers,and actively generate and interact with knowledge;(2)user-generated content in social Q&A sites creates a complex knowledge network through question tags or semantics,and the highly aggregated knowledge unit is the skeleton structure of the knowledge network;(3)the analysis results of the knowledge network based on the text topic have good accuracy,which can not only extract the knowledge units added and defined by users,but also extract other semantics in the text and expand the boundaries of the knowledge network;(4)in the empirical dataset,the uncertainty distribution of the knowledge network converges rapidly,and its change will become very slow after converging to a stable distribution,indicating that the network structure tends to be stable from the initial chaos and disorder.