Study On Topic Propagated Path Reconstruction Technology Based On Time And Content

With the development of the Internet as a ubiquitous computing and communication platform,social networking technology is integrated into the information society.Users can express themselves and make friends in these social networking sites,such as blogs,firums,WeChat.Therefore,hundreds of millions of information generated everyday.Studying how to reconstruct the topic propagation path have great significant in research and network security.It can be applied to trace to the source,to show the information dissemination,and guide the propagation analysis.In conclusion,the path reconstruction of topic propagation is essential in propagation analysis,public opinion monitoring and other areas.To solve the problem,this paper proposes a topic propagated reconstruction method based on the time and content.Firstly,we use the topic detection method to cluster the weibo into different group.Secondly,we study the influence of social relationship,text similarity and time in topic diffusion,and build a modal for transmission probability.And then,we obtain the information source node for each node with the method of back propagation to construct the propagated path map in social network of a topic.Through the information propagated path map,we can get information about the aggregation trend of the opinion information,the real propagation direction and the prediction of the similar topic.First of all,we determine the impact of several factors through the analysis of factors affecting information spread.Secondly,based on the experimental background and the characteristics of text transmission,we propose an improved weighted similarity calculation method of weighted Jaccard by analyzing the existing text similarity algorithm.Then,through the study of the distribution of time interval under different trends of the development of topics,we study the influence of time in topic diffusion.After which,we design the topic propagation probability calculation method to find the implicit propagation path between users.Finally,combined this method with text processing,the historical propagation path graph of the topic is obtained.