摘要
随着互联网的快速发展,从海量信息中获取感兴趣的信息越来越困难。推荐系统正是解决这一难题最热门的技术之一。数据稀疏性问题是当前推荐系统所面临的主要问题之一。为了缓解数据稀疏性的问题,本文借助社交网络,提出了一种融合用户社交网络的推荐算法,将用户在社交网络中的亲密度引入推荐系统。在实验部分,本文采用百度电影推荐算法创新大赛的数据集,设计实验验证了提出算法的有效性。实验结果表明,本文提出的算法能够有效地缓解传统协同过滤算法面临的数据稀疏性问题,明显提高预测的准确性。
With the rapid development of the Internet, people are in a time of explosive growth of information. It＆#39;s more and more dififcult for people to ifnd out the needed information. Recommendation system is currently the most popular and effective techniques to solve this problem. The paper proposes one recommendation algorithm based on social network. With the help of the social network, the algorithm introduces the social intimate degree into the recommendation system. The paper designs experiment using the Baidu movie recommendation algorithm contest＆#39;s dataset. Experiments show that social and rating cross-domain recommendation system can improve prediction accuracy of the system.
出处
《软件》
2013年第12期41-45,共5页
Software
关键词
数据挖掘
个性化推荐
社交网络
协同过滤
Date mining
Recommendation System
Social Networking
Collaborative Filtering
