欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
在线社交网络成为人们网络生活的最主要平台,人们基于兴趣偏好等原因集聚形成各个网络社区,共同参与感兴趣话题的讨论,表达自己的观点和看法,寻找感兴趣的内容,因此识别在线社交网络中用户的兴趣偏好,具有重要意义.本文首先分析用户在社交网络中的行为,提出从用户发布信息、基于共同参与话题的社交关系中寻找相似兴趣最近邻,和再考虑用户在社交网络中影响力的相似兴趣最近邻,三种获取用户兴趣偏好的方法.最后利用百度贴吧数据集实验,比较三种方法的使用效果.实验表明考虑用户影响力的最近邻方法获取用户兴趣偏好的方法效果最好,而且这种方法不需要利用用户本身信息,仅仅只需要通过其最近邻用户就可以获得更加准确的用户兴趣偏好.
英文摘要：
Online social networks （OSNs） emerged as a popular platform for people＇s online life, people gather into virtual communities based on their preferences. They involved in topic discussion, expressing their views and opinions, and searching for content that they are interest in. So discover the interests and preference of people in the community is of great significance. In this paper we studied users＇ behavior in social networks, and presented 3 methods of getting users＇ preference from user-generated content （UGC） , from the nearest neighbors and from the influential nearest neighbors. Finally, we compare the application results of the 3 methods by the case of Baidu PostBar. The experiment showed that the method of considering the user influence of nearest neighbors is the best in our research. Use this method does not require users＇ information themselves, only need to obtain information from the nearest influential neighbors.
同期刊论文项目
同项目期刊论文
期刊信息
位置：