A Simple and Heuristic model of Tag Recommendation

摘要：

Compared to the high computational complexity of many tag recommenders, a simple and heuristic approach of tag recommendation is proposed, based on tag user and item tag co-occurrences in parallel. Firstly, we use aspect model PLSA to set up a probabilistic model. We find that the probability of the recommended tags to an item for a specific user is determined by two factors: the preferences in choosing tags for the user and the tags reflecting the feature of the item. Then we immerge the two factors into a unified representation. The experiments show that our approach not only has better reliability and precision, but also is very simple and more practical than other algorithms.

关键词： social tagging topic probabilisic model personlized tag recommendation

作者: Dihua Xu Zhijian Wang Liping He Weidong Huang

作者单位: College of Computer and Information, HoHai University College of Computer, Nanjing University of Pos College of Computer and Information, HoHai University Nanjing, China College of Education Science and Technology, Nanjing University of Posts Telecommunications Nanjing, College of Economics and Management, Nanjing University of Posts Telecommunications Nanjing, China

会议类型: 国际会议

会议名称: 2011 Fourth International Symposium on Computational Interlligence and Design 第四届计算智能与设计国际会议 ISCID 2011

会议地点: 杭州

会议语种:英文

页码: 542-545

在线出版日期: 2011-10-28（万方平台首次上网日期，不代表论文的发表时间）