基于时效性的冷启动解决算法

Timeliness-Based Algorithm for Cold Start

在推荐系统研究领域,协同过滤推荐技术是一种重要的技术方法,但新用户和新项目等冷启动问题是该技术方法所面对的一个重要问题。为解决新用户冷启动问题,充分利用评分数据上下文信息,提出一种基于项目时效性模型的解决算法,把时效性高的项目推荐给刚加入系统的新用户,从而缓解新用户冷启动问题。实验结果验证所提出的算法在保证推荐精度的情况下能为新用户产生有效的推荐。

Collaborative filtering recommendation technology is the most important technology in recommender systems, but the technology is facing new users and new items cold start problem. To solve the new users cold start problem, proposes a solution algorithm based on item timeliness model by making full use of the context information of rating data, and recommends high timeliness items for new users. The experimental results verify that the algorithm proposed produces effective recommendation for new users.