【摘要】 机会网络的节点相遇形成了社会关系.现有的机会网络社团划分都是以确定的社会关系为输入.然而,由于节点之间的社会关系由相遇和通信共同决定,这使节点间的社会关系存在不确定性.为了研究机会网络社会关系及其社团,建立了机会网络的不确定社会关系模型,并基于该社会关系模型进行社团划分.首先,根据节点相遇、通信记录,构建机会网络的不确定社会关系模型;其次,提出了社团概率密度,并根据社团概率密度提出了改进的K派系过滤算法,该算法能够对不确定的社会关系进行社团划分.实验结果表明,基于社团概率密度的K派系过滤算法能够得到较好的社团划分结果.更多还原

【Abstract】 Opportunistic network nodes encounter and form a type of social relationship. All of existing opportunistic networks use definite social networks as the input. However,the social relationship of nodes is determined by encounter and communication between nodes,which leads to their uncertain relationship. To examine the social relationship of opportunistic networks and communities,we developed a model for uncertainty social relationship of opportunistic networks,and then detected communities based on our model.Firstly,we developed this model according to the records of encounter and communication between nodes. Next,we proposed the concept of community expect and use it to improve K-CLIQUE which can be employed to detect communities among uncertainty social relationships. Finally,our experiments indicated that the K-CLIQUE based on community expect can get a better result of community detection.更多还原