The Study Of Dynamic Micro-blog Community Based On Structure And Gravity Bin-cohesion

Thanks to the continuous progress of human society and the rapid development of the Internet,People's Daily life and work are inseparable from the Internet.Due to the convenience brought by the Internet,people can do many things in their daily life through the Internet directly,such as paying,shopping,making friends and so on.Sina Micro-blog,Twitter and other social network platforms are important ways for people to communicate with each other and access information.People can use these ways to share their daily life,express their opinions and so on.Due to the growing number of users in social network platforms,massive amounts of data are generated every day,which allows us to naturally enter into the big data era.It is very important that how to use massive amounts of data to solve the existing problems in our daily life.Micro-blog is an open and real-time online social network platform based on the users and their relations,and allows people to post daily events,share personal feelings and more.Among the large number of Micro-blog users,some Micro-blog users actually have similar interests and hobbies.Therefore,it will be very significant to gather Micro-blog users who have similar interests and hobbies,such as advertising accurately,recommend friends.Community refers to a group of nodes,and community detection is the detection of such a group with a special relationship.Prior approaches for community detection are analyzed from the community structure.The community is not static over time.Therefore,the study of community evolution is also indispensable.The existing methods of community evolution mainly study the explicit evolution of the community,such as "split","merge" and "shrink".For the Micro-blog social network,we can not only analyze the community from the community structure,but also extract the Micro-blog user interest features from the Micro-blog content.As for the evolution of Micro-blog community,some existing community evolution methods only analyze the explicit evolution behavior of the community.However,Micro-blog communities still have implicit evolution behavior.Based on the existing shortcomings of community detection methods and community evolution methods,we will conduct our research work.The main contents of this research include the following:1.Giving the definition of the Micro-blog social network based on the Micro-blog data with time series,including the definition of the static Micro-blog social network,the definition of the dynamic Micro-blog social network,the definition of the Micro-blog community,the definition of community evolution as well as the definition of community evolution behavior.Using Natural Language Processing Technology to extract Micro-blog user interest features from Micro-blog Data Sets.Reconstruct the gravity relations in complex networks,and combine the extracted interest features of Micro-blog users to construct thegravity relations in the Micro-blog social network.According to the new definition of Micro-blog social network and the gravity relations,Micro-blog social network map is constructed.2.Using the random walk method combined with game theory to detect the Micro-blog user's gravity tendency.Based on constructed the relations in Micro-blog social network map and the discovered gravity tendency of the Micro-blog users,we find out the basic nodes in the Micro-blog social network and propose the Micro-blog community detection algorithm to detect the community in the Micro-blog social network.3.Based on the community at time t,we find out the changes of the Micro-blog social network in adjacent time series through the Micro-blog data at time t+1.According to the Micro-blog community at time t and the Micro-blog users who have changed in adjacent time series,a Micro-blog community evolution algorithm is proposed to detect community at time t+1.We define the probability of mutual transformation between communities at time t and time t+1 and propose the Micro-blog community evolution behavior extracting algorithm to extract the social evolution behavior in Micro-blog social network.4.Conducting experiment to evaluate the algorithms we proposed.The experiment includes three parts: the comparison and evaluation of the Micro-blog community detection algorithm,the comparison and evaluation of the Micro-blog community evolution algorithm,and the research on the optimal range of the parameters involved in the algorithms.The results obtained in the experiment are represented by tables and graphs,and are analyzed.The experiment results show that the proposed algorithms has good performance with other benchmarking methods in structure,and has better performance than other benchmarking methods in gravity.