Inferring User Interests in Microblogging Social Networks: A Survey

摘要

With the popularity of microblogging services such as Twitter in recentyears, an increasing number of users use these services in their daily lives.The huge volume of information generated by users raises new opportunities invarious applications and areas. Inferring user interests plays a significantrole in providing personalized recommendations on microblogging services, andthird-party applications providing social logins via these services, especiallyin cold-start situations. In this survey, we review user modeling strategieswith respect to inferring user interests in previous studies. To this end, wefocus on four dimensions of inferring user interest profiles: (1) datacollection, (2) representation of user interest profiles, (3) construction andenhancement of user interest profiles, and (4) the evaluation of theconstructed profiles. Through this survey, we aim to provide an overview ofstate-of-the-art user modeling strategies for inferring user interest profileson microblogging social networks with respect to the four dimensions. For eachdimension, we review and summarize previous studies based on specifiedcriteria. Finally, we discuss some challenges and opportunities for future workin this research domain.