论文部分内容阅读
根据社会网络中人们的行动决策会受到周围人群影响的特征,在Web社会网络环境中提出了一个引入群体凝聚度的信任模型GC-Trust.模型首先对节点进行凝聚群的划分,其次从两个层面研究信任度,一是凝聚群之间的信任度,二是凝聚群与单个节点之间的信任度,再利用节点所在凝聚群内的群凝聚度以及节点在自身凝聚群中的影响力作为不同信任度之间的权重将他们进行综合,从而综合形成节点的凝聚信任度,以便选择合适的对象.实验表明GC-Trust模型与Tidal Trust模型以及基于同质度的信任模型相比,在凝聚群规模相对平衡且凝聚度高的环境中具有更高信任预测准确度.
According to the characteristics of the people in the social network, the decision-making of the people will be influenced by the surrounding people. In the Web social network environment, a trust model GC-Trust which introduces the cohesion of the group is proposed. The model first divides the nodes into two groups The first is the trust between agglomeration groups. The second is the trust between agglomeration groups and individual nodes, and then uses the agglomeration degree of agglomeration within the agglomeration group where nodes are located and the influence of nodes in their own agglomeration group The weights between different trust levels combine them to synthesize the node’s cohesion trust in order to select the appropriate object.Experiments show that compared with the Tidal Trust model and the homogeneity-based trust model, In a relatively balanced and cohesive environment, the cluster has a higher degree of confidence in forecasting confidence.