Personalized On-line Service of Particle Swarm Optimization Cluster Analysis Algorithm

To provide accurate and personalized information service, a k-median cluster analysis of parallel particle swarm optimization (PSO) and simulated annealing (SA) algorithm was proposed to cluster the information resource in the network. The algorithm integrated the fast search optimum ability of parallel PSO with probability jump property of SA. It can maintain the individual diversity and avoid the degenerate phenomenon. To aggregate information, the cluster analysis algorithm adopted reiterative contrasting of user´s multiple interesting vector with information resource, which was created by k-median cluster analysis according to the input search key words and their weight. If the process of information clustering was accomplished, a information catalog would be validated. Based on this catalog, integrated with on-line service technology and customer demand, the information could be piloted to perform second match, so as to guarantee the accuracy of personalized information service