从复杂性科学角度总结了社会化标签系统的结构、演化和功能问题.包括多类异质性节点和超图模型的网络结构;基于标注行为的网络演化模型;基于标签的个性化推荐系统.系统地总结和比较了当前几种代表性的模型和推荐算法,并指出各方法的优缺点和适用范围.有助于更深层次理解和解决社会化标签系统中的理论和应用问题.
The recent research progress on understanding the structure, dynamics and function of social tagging systems was summarized. In particular,the following was emphasized:network structures of heterogeneous nodes and hypergraph; tagging behavior based evolving models; tag-aware recommender systems. The representive models and algorithms were systematically studied and compared, and advantages and disadvantages of each, as well as their applicable areas, were pointed out respectively. This survey could shed some lights on deep understanding on both theoretical and application problems of social tagging systems.