Time-factor-based group recommender algorithm on social networks

• 作 者: ZUFENGZHONG;YAOQINGDUAN;
• 机 构: Department of Information Management,Central China Normal University,Wuhan 430079,China
• 出 处: 第十三届海峡两岸图书资讯学学术研讨会
• 发表日期: 2016年期

　　The existing recommender systems are most aimed at single users,rather than groups.A group in social networks is characterized by complexity and diversity,so the traditional group recommender algorithms(GRAs)are unable to adapt to social networks.Therefore,we proposed a time-factor-based hybrid GRA for personalized Weibo recommendation on social networks.The new algorithm comprehensively considers the between-user interaction and determines the importance degree of a microblog through extraction of current context information.Moreover,it considers the between-group relation structure and interests also change with time.Here the real data from Sina Weibo were used in test and evaluation.The new algorithm outperforms the existing algorithms in Weibo group recommendation.