融合云环境用户情境兴趣的移动SNS信任推荐模型  

Trust Recommender Model for Mobile Social Network Integrating Cloud Environment User's Situational Interest

作　　者：宓翠 陈晶 苏妍螈 张亚明 刘海鸥 

机构地区：[1]燕山大学经济管理学院,互联网+与产业发展研究中心,河北秦皇岛066004

出　　处：《小型微型计算机系统》2018年第3期484-489,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(71271186,71671159)资助;教育部人文社科基金项目(17YJCZHl09)资助;河北省自然科学基金项目(G2017203319,G2016203220,G2015203380)资助.

摘　　要：传统协同过滤(Collaborative Filtering,简称CF)推荐面临的数据稀疏性问题是当前移动社交网络服务(Social Networking Services,简称SNS)个性化推荐亟待解决的问题.在回顾相关文献的基础上,提出融合云环境用户情境兴趣的移动SNS信任推荐模型,融合移动SNS情境相似度矩阵与信任度矩阵进行CF推荐,基于此缓解数据稀疏性导致的推荐性能下降问题,同时采用MapReduce并行处理方式提高大规模复杂社交网络的并行挖掘性能.实验结果表明,融合云环境用户情境兴趣的移动SNS信任推荐模型较好地将情境相似度与信任度融入到CF推荐过程中,缓解了由于数据稀疏与信任稀疏导致的推荐精度下降问题,MapReduce化的并行处理方法也有效提升了大规模复杂社交网络的并行推荐性能.One of the key problems facing mobile SNS personalized recommendation is the data sparse and cold start problem. By re- viewing the research progress, we proposed a trust recommender model for mobile SNS based on contextual interest in cloud environ- ment. Based on combining user contextual interest similarity and trust similarity together, the model could effectively alleviate the sparse data problems. Finally, we used MapReduce to improve the large-scale complex social network parallel performance. The ex- perimental results showed that the method smoothly introduced the interest similarity and social trust networks into traditional CF, and solved the recommendation accuracy due to sparse data problems effectively. Besides, the parallel method effectively solved the rec- ommend problem of large-scale complex social network and improved the real-time recommendation performance.

关 键 词：云环境 情境兴趣 移动SNS 信任推荐模型 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]