所以,我想实现一种混合协同过滤算法(Full algorithm here)。该算法有一个评级表
评分:=(id,user_id,item_id,评分)
此表基本上记录了项目的用户评分。我从未实现推荐引擎,但基本上我想计算:
两个用户的常用商品Formula and info can be found here,can also be seen here, page 501, marked as (1)
熵,超熵和期望值Formulas can be seen here on page 502, marked as (6)