Combining long-term and short-term user interest for personalized hashtag recommendation  

作　　者：Jianjun YU Tongyu ZHU 

机构地区：[1]Computer Network Information Center, Chinese Academy of Sciences, Beijing 100190, China [2]State Key Lab of Software Development Environment, B eihang University, Beijing 100190, China

出　　处：《Frontiers of Computer Science》2015年第4期608-622,共15页中国计算机科学前沿（英文版）

摘　　要：Hashtags, terms prefixed by a hash-symbol #, are widely used and inserted anywhere within short messages （tweets） on micro-blogging systems as they present rich sen- timent information on topics that people are interested in. In this paper, we focus on the problem of hashtag recommenda- tion considering their personalized and temporal aspects. As far as we know, this is the first work addressing this issue spe- cially to recommend personalized hashtags combining long- term and short-term user interest. We introduce three features to capture personal and temporal user interest： 1） hashtag textual information; 2） user behavior; and 3） time. We of- fer two recommendation models for comparison： a linear- combined model, and an enhanced session-based temporal graph （STG） model, Topic-STG, considering the features to learn user preferences and subsequently recommend person- alized hashtags. Experiments on two real tweet datasets illus- trate the effectiveness of the proposed models and algorithms.Hashtags, terms prefixed by a hash-symbol #, are widely used and inserted anywhere within short messages （tweets） on micro-blogging systems as they present rich sen- timent information on topics that people are interested in. In this paper, we focus on the problem of hashtag recommenda- tion considering their personalized and temporal aspects. As far as we know, this is the first work addressing this issue spe- cially to recommend personalized hashtags combining long- term and short-term user interest. We introduce three features to capture personal and temporal user interest： 1） hashtag textual information; 2） user behavior; and 3） time. We of- fer two recommendation models for comparison： a linear- combined model, and an enhanced session-based temporal graph （STG） model, Topic-STG, considering the features to learn user preferences and subsequently recommend person- alized hashtags. Experiments on two real tweet datasets illus- trate the effectiveness of the proposed models and algorithms.

关 键 词：RECOMMENDATION hashtag time-sensitive userinterest 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术] TV698.11[自动化与计算机技术—计算机科学与技术]