一种基于信任度和研究兴趣的学者推荐方法  

A Recommendation Method for Scholars Based on Trust and Research Interests

作　　者：孙赛美 林雪琴 彭博 李春英 汤庸[1] SUN Saimei;LIN Xueqin;PENG Bo;LI Chunying;TANG Yong(School of Computer Science,South China Normal University,Guangzhou 510631;School of Computer Science,Guangdong Polytechnic Normal University,Guangzhou 510655)

机构地区：[1]华南师范大学计算机学院,广州510631 [2]广东技术师范大学计算机科学学院,广州510665

出　　处：《计算机与数字工程》2019年第3期608-615,共8页Computer & Digital Engineering

基　　金：国家自然科学基金(编号:61772211);广州市科技计划项目(编号:201604046017;201704020203)资助

摘　　要：学术社交网络用户的爆炸式增长,使得用户难以发现自己感兴趣的相似学者,对此论文提出一种融合信任度和研究兴趣相似度的学者推荐方法。首先通过用户在社交网络的交互行为计算用户间的信任度来衡量好友关系的真实强度,再利用三度影响力理论扩展用户的潜在好友集合,然后再利用LDA信任度和研究兴趣相似度进行最终推荐。在学者网数据集上的验证表明,该方法有效提高了学者推荐的效果。As the exponential growth of the number of users in academic social networks,it makes it difficult for users to find similar scholars whom users are interested in. Thus,a recommendation method for scholars based on trust and research interests is proposed. Firstly,the trust between users is calculated to measuring the valid relation strength by using the interaction behavior of the user in the social network. Secondly,the three degrees of influence rule is utilized to expand the users' potential friend set. Third. ly,similarity of research interests among users is calculated by the LDA topic model. Finally,the trust and similarity are integrated as the comprehensive similarity by the weighted mean method to recommend users. The experimental results on datasets of SCHO. LAT demonstrate that the method can effectively improve the performance of scholar recommendation in terms of precision and re. call.

关 键 词：学术社交网络 学者推荐 信任度 三度影响力 LDA主题模型 

分 类 号：TN711[电子电信—电路与系统]