摘要
针对社交网络服务中汇聚的大量带有地理和社交属性的数据,提出一种改进的综合考虑用户位置和好友关系的个性化位置Top-k查询方法。首先,在传统Top-k空间查询方法的基础上,将用户的好友关系及好友签到统计信息引入评分函数,以便对查询结果进行个性化排序。然后,改进IR-tree索引结构以支持对位置信息和社交关系的混合索引。最后,在查询过程中采用基于优先级队列的最佳优先遍历方法进行剪枝,从而减少搜索空间。实验表明,改进的评分函数、索引结构和遍历方法能够实现社交网络环境中个性化且高效的Top-k查询。
Regarding a lot of geographical and social data gathered in social networking services, a personalized location Top-k query method which takes into account both location and social relationship is presented. First, a personalized rank- ing function is presented based on the traditional Top-k spatial query method. It combines the user~ social relations with the statisties information of friends＇chick-ins. Then, the IR-tree index is improved to support the combined index on location information and social relations. Finally, the best-first traversal methods based on priority queue is adopted for pruning to reduce the search space. The experimental results showed that the improved ranking function, index structure and traversal method could implement both personalized query and efficient query in social networks.
出处
《重庆邮电大学学报（自然科学版）》
CSCD
北大核心
2013年第5期644-650,共7页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
重庆市自然科学基金项目(cstc2012jjA40014)
重庆邮电大学博士启动基金项目(A2012-34)~~
