在个性化推荐系统中,用户模型是推荐系统的主要依据,模型的好坏直接影响推荐系统的质量.但是在实际的应用中,用户兴趣会随环境和时间发生变化,有可能推荐的商品已不满足用户当前的兴趣需求.本文提出了一种记录用户当前兴趣模型自我更新的算法,通过对用户历史访问数据建立用户兴趣模型,以时间为主线,统计用户当前兴趣.本文以Movielens电影数据集为数据源,在设计的个人推荐系统中应用基于用户行为反馈的用户兴趣模型更新算法,有效缓解了用户兴趣偏移的问题,提高了推荐质量.%In the personalized recommendation system, the user model is the main basis of the recommended system, the model directly affects the quality of the recommended system.However, in practical applications, the user interest will change with the environment and time, it is possible to recommend the goods do not meet the user′s current interest needs.This paper presents an algorithm to record the user′s current interest model self-renewal.By setting the user interest model to the user historical access data, the time is the main line, and the user′s current interest is counted.In this paper, the Movielens film data set is used as the data source, and the user interest model updating algorithm based on user behavior feedback is applied in the design of the personal recommendation system, which effectively alleviates the problem of user interest offset and improves the recommendation quality.
展开▼
