针对协同过滤系统冷启动问题的混合优化算法

Hybrid optimization algorithm for cold start problem of collaborative filtering system

为更好地解决协同过滤推荐系统存在的冷启动问题,依据对现有信任网络和用户标签行为的概念研究,明确其算法思想,对原有协同过滤系统算法做出优化。通过对原有信任网络算法和用户标签化行为算法逐一进行优化,并凭借一定的手段有效地使二者结合,建立了一个效能较好的推荐系统算法。经试验验证该算法在一定程度上较原有算法有了更好的改进,能更好地解决协同过滤冷启动问题。对协同过滤推荐系统的冷启动问题的研究提供了新的思路和方法。

in order to better solve the cold start problem of collaborative filtering recommendation system, based on the concept research of existing trust network and user tag behavior, the algorithm idea was defined, and the algorithm of the original collaborative filtering system was optimized. By optimizing the original trust network algorithm and user labeling behavior algorithm one by one, and combining them effectively by certain means, a good recommendation system algorithm is established. Experimental results show that this algorithm is better than the original algorithm and can solve the cold start problem of collaborative filtering. It provides a new idea and method for the research of cold start of collaborative filtering recommendation system.