针对社区网站中通过衡量用户静态信息的一致性和共同好友数量，忽视其动态信息以及动态信息变化过程实现好友推荐这一问题，提出基于兴趣集、兴趣度持续时间、兴趣集序列构造性格模型，比较用户性格相似度实现性格相似的判断。实验结果显示，该模型实现好友推荐的效果良好。%In view of SNS, the way to achieve friends recommended by measures the consistency of user’s static information or the number of common friends. But it ignores dynamic information and the change process of it. For this problem, this paper based on interest sets, the duration of interest and the interest sequence to structure a character model. Compares the similarity to achieve character similarity judgment. The results show that it is a good model to get friends.
展开▼
