基于Python的推荐系统相似性分析和协同过滤

Similarity Analysis and Collaborative Filtering in Recommendation System Based on Python

推荐系统是根据用户在电子商务网站上的购买和浏览记录,将用户感兴趣的物品主动推荐给用户。有两种主要的推荐方法,一种是基于用户相似的推荐,一种是基于项目相似的推荐。文中介绍了计算用户或项目的相似程度的常用方法,诸如欧式距离、曼哈顿距离、皮尔逊相关系数、余弦相关性等算法。在本章的最后,还给出了一个基于项目协同过滤的推荐系统的Python分析和计算。

According to user’s purchasing history and browsing records on e-commerce web sites, the recommender system will recommend the user’s interested items to the user. There are two main recommendations, one based on the similarity of the user, the other based on the similarity of the item. Common algorithms of calculating the similarity of the user or item, such as Euclidean distance, Manhattan distance, Pearson correlation coefficient and Cosine similarity is introduced. In the end of this paper, the analysis and calculation of an item-based collaborative filtering recommendation system with Python is also given.