Relationship Mining And Optimization Design In Wireless Social Network

Wireless social network,as a form of social organization based on node not group,contains all kinds of social relationships,including friend relationship,schoolmate relationship and family relationship and so on.Through these relationships,wireless social network ties the people from casual acquaintance to cohesive family in series.With industrialization,urbanization and the rise of new communication and Internet technologies,society is becoming more and more networking,the research about wireless social network is becoming more and more important as well.In the meantime,wireless social network,as the upper layer,is relying on the underlying technology network.Through the research about wireless social network,we can provide new directions and ideas for the design and optimization of underlying technology network,thereby improving the network's performance.The research about wireless social network includes two aspects.First,we study the network structure of wireless social network based on the graph theory,such as degree distribution,clustering,small world phenomenon.Second,we analysis the user behavior in wireless social network based on the data mining.The research in this paper is based on these two aspects.Firstly,the caching-enabled base station of the wireless social network is picked as an example,which has the issue of high network delay.We introduce one of the network structure,small world phenomenon,into the caching-enabled base station.And research analysis prove that constructing the small world model in caching-enabled base station can reduce the average network delay in an efficient way.Also,a delay-optimized deterministic shortcut addition strategy is proposed.The numerical simulation result shows the efficiency of small world model in reducing the average network delay.Meanwhile,compared with other small world construction strategies,the proposed delay-optimized shortcut addition strategy can maximum the performance in decreasing the average network delay.Secondly,the massive online open courses which is formed by wireless social network and online education is playing a more and more important role in education.However,massive online open courses has to face the issue that the forum information is so redundancy and complex that user can't get effective information from it.In order to solve this problem,a recommendation system is proposed based on data mining technology to recommend users the forum threads which is relevant to the courses.Also,a classifier iterative optimization algorithm based on correlation weights is proposed to improve the accuracy of classifier.Results improve the efficiency of recommendation system,and showing that the proposed optimization algorithm can improve the precision of the classifier in recommendation system.