Detecting community structure in trust relationship networks

In crowd-computing service, untrusted service nodes threaten the user security and service quality. To find the trusted node sets (community structure) in trust relationship networks, we propose a novel scheme named community structure detecting algorithm in trust relationship networks (CDATN). In this scheme, we first introduce the factors of weight and direction to build a directed-weighted model for the trust relationship network. Then we define the vertex similarity index and design the objective function to control the clustering process. Finally, the service nodes can be eventually divided into different trusted node sets, which provide a reference for the service nodes selection. Experimental results show that the scheme can effectively detect and identify the trusted node sets in the trust relationship network, and presents a better performance in contrast to the existing algorithm.