Study of users aggregation algorithm based on static social tags

Nowadays the core of social aggregation service is still about polymerization and recommendation of content. Yet, the mining of the user association behind the shared information is unsatisfactory. Based on static social tags, a new algorithm clustering users is proposed, which clusters users by their similarity. Firstly, relevant concepts are defined to be used in the algorithm. And then, by analyzing existing clustering algorithms, K-means is chosen to be improved. The improved K-means algorithm is proposed and the basic flow of the algorithm is represented. Next, data from a Folksonomy site is used to conduct our empirical study, comparing the clustering effects of the two algorithms by experiment. As a result, the improved K-means algorithm shows better clustering effects than that of original algorithm and is suitable for clustering users.