摘要
该文在分析推荐系统和进化计算理论的基础上,提出一种多目标优化思路,给出一种基于进化多目标优化的推荐系统算法。该算法同时考虑推荐的精确度和推荐的新颖度,既要保证精确率又要尽可能地推荐新的物品给用户,算法力求在两者之间得到一种平衡。该文给出算法的设计思想和算法流程,并对算法进行了模拟数据的测试。
Based on the theoretical analysis and recommendation system evolution,this paper has proposed a multi-objective optimization idea and an evolutionary multi-objective optimization based recommendation algorithm is proposed.This algorithm simultaneously considers the recommendation precision and novelty,it not only preserves precision but also recommend new items to user,it makes effort to obtain the tradeoff between these two objectives.This paper presents the design of algorithms and algorithmic thinking processes,and tests the algorithm with simulation data.
作者
樊鸿
FAN Hong (School of computer,Xi'an University of Electronic Science and Technology, Xi'an 710000, China)
出处
《电脑知识与技术》
2014年第4期2342-2346,共5页
Computer Knowledge and Technology
关键词
推荐系统
进化优化
recommendation systems
evolutionary optimization