Study Of Micro-blog User Relationship Network Structure Based On Complex Network

With the development of Internet, many different information networks, likeMicro-Blog, have emerged. As a social network and a new media, Micro-Blog is oneof the most popular Internet services because it is easy to be used for people and theinformation in it can be spread very rapidly. In one Micro-Blog network, the basicrelationship between users is following relation which can be established easily. Suchfollowing relationships in one Micro-Blog network are very complex. The research onusers’ relationships is very important with extensive application prospect. Theresearch achievements can be applied into public sentiment monitoring, informationrecommending, network marketing.In this thesis, the "sina Weibo "(Sina Micro-Blog) is used as an example forexploring the general rules of users’ following relationships by applying complexnetwork analysis methods. Firstly, the scale-free and small world properties of theWeibo following relationships network are studied to reveal that the network hasobvious complex network features.Secondly, the K-core decomposition operation is performed on the Micro-Blogfollowing relationships network to obtain a core users’ relationship network. Byanalyzing the K-core network’ basic parameters, following ratio, centrality and degreecorrelation, a conclusion can be presented that the K-core network has the features ofboth an online community network and a realistic society network. At the same time,we can know that the network structure presents obvious community characteristicsand users in a community establish the network using relationships by communitydetecting on the K-core network and analysis of the node centrality, reciprocity andthe role of brokerage. This provides a new way to improve information exchangesystem of a Micro-Blog community.Thirdly, a comment relationship network between users is constructed and itscomplex network features are studied. The experimental results show that thecommunications among users in a same community have correlation with the socialattribute. The community with the most active communications is the one constructedby the following relationship between classmates or between teachers and students.And the community with the sparsest communications is built by the relationships between advertising marketing. Through the analysis of core users, we can know thatthe core users used to transfer information in a following relationships network aredifferent from those in a comment relationship network, which reveal the features ofinformation communication in a Micro-Blog community.Finally, a summarization on the main work in this thesis is presented, and someunsolved problems on Micro-Blog network structure by using complex networkanalysis methods are also described, which can be considered as our main researchwork in the future.