Research On The Key Technology Of Situation Awareness In Network Group Event

The internet is gradually evolving into a ubiquitous computing platform and information dissemination platform.The emergence and rapid development of online social networks,BBS and other social media have changed our understanding of internet from simple browsing and searching to the construction and maintenance of online social relationship.Users aggregate and interact in these platforms,in where the information based on social relationships is created,communicated and disseminated.Started from topic evolution and group role analysis,the thesis studies the dissemination and source tracing of online group events.This thesis is mainly focused on the following two aspects:1.In the analysis of online group events evolution,based on user behavior and relationship,we propose a situation awareness model for social events to predict user participation behavior and event development trends.Firstly,for the complex factors of user behavior,three dynamic influence factor functions are defined,including individual,peer and community influence.These functions take timeliness into account using a time discretization method.Secondly,to dynamically discover the patterns of individual and group behavior within a social topic,a situation awareness model is proposed associated with the basic concepts of random field and Markov property of information diffusion.The experimental results show that the model can not only dynamically predict the individual behavior,but also grasp the development trends of event.2.In the aspect of group role analysis,an information tracing model is proposed leveraging the participated users and their relationship.This model discovers the publisher and important users of social event.Firstly,the development trend of online event is studied,and participated users of the event is extracted.Secondly,we build the structure network by relationship among the participated users.Then,the contribution rate of participated users to the development of the event are initialized,and the PageRank algorithm is introduced to construct information tracing model.Finally,according to the contribution rate calculated by the model,the Topk users are selected as the publisher and important users for this event.Experimental results show that our model can discover the publisher and important users of this event.