推荐系统研究进展与应用

Progress and Application of Recommendation System

推荐系统一直以来都是很活跃的研究方向。回顾了近5年来推荐系统涉及的数据集、算法、效果评估的研究发展状况,给出了推荐系统在互联网应用场景中的系统架构,从工业知识与经验以及工业生产工艺与工艺参数这两方面对推荐系统工业生产中的最新应用做了介绍。分析了推荐系统中“信息茧房”问题产生的原因,展望了未来推荐系统在“信息茧房”、可解释性、隐私性等方面的研究。

Recommendation system has always been a very active research direction. This paper reviews the research and development of datasets, algorithms and effect evaluation involved in the recommendation system in recent 5 years, gives the system architecture of the recommendation system in the Internet application scenario, and introduces the latest application of the recommendation system in industrial production from two aspects: industrial knowledge and experience, industrial production process and process parameters. This paper analyzes the causes of the problem of “information cocoon room” in the recommendation system, and looks forward to the research on“information cocoon room”, interpretability and privacy of the recommendation system in the future.