用用户来表示隐特征：
——用用户个体解释单个隐因子。
1） （利用分解后的物品矩阵）分解之后，计算用户的评分向量和隐特征关于所有物品的分解向量（物品隐因子矩阵中以隐因子为索引取行或者列），然后，用距离最近的用户来解释因向量。
2）（利用分解后的用户矩阵）通过观察分解后的用户矩阵的形式，如果是类似于一个元素特很大，其他元素解决于0的话，那么就可以用这个用户表示该对应的隐因子。（We can say that if the vector w_l of user u_l has the following form: the k^th value of the vector is high and other values are close to 0, this means that this user is highly related to feature f_k . Thus, this latent feature f_k may be interpreted as user u_l.）
3）同时利用以上两种方式：如果某个用户同时利用以上两种方式确定为representative user，那么这个用户更应该用来解释对应的隐因子
用representative user来解决cold-start problem
要求representative user对new items做出评分，然后利用向量乘积和求解方程式的方式，求出new items的隐因子分解向量，然后就可以预测其他用户关于new items的评分了
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| Can Latent Features be Interpreted as Users in Matrix Factorization-based Recommender Systems? | |||||||||||
| 文献作者 | Armelle Brun; Marharyta Aleksandrova | ||||||||||
| 文献发表年限 | 2014 | ||||||||||
| 文献关键字 | |||||||||||
| representative users; interpretable; cold-start problem; | |||||||||||
| 摘要描述 | |||||||||||
| Matrix factorization has proven to be one of the most accurate recommendation approach. However, it faces one main shortcoming: the latent features that result from factoriza- tion, and that represent the underlying relation between users and items, are not directly interpretable. Some works focused on their interpretation, particularly with non-negative matrix factorization. In these works, features are viewed as groups of users, groups of items or as attributes of items, but such interpretations require human expertise. In this paper, we propose to interpret features as users, referred to as representative users. This interpretation relies on the study of the matrices that result from the factorization and on their link with the original rating matrix. Such an interpretation is not only performed automatically, as it does not require any human expertise, but it helps also to explain the recommendations made to users. In addition, we see it as a way to alleviate the new item cold-start problem, without requiring any information about the content of the items. The experiments conducted on several benchmark datasets confirm that the features discovered by a non-negative matrix factorization can be actually interpreted as users and that the representative users (the interpretations of the features), are a reliable source of information that allows to accurately estimate ratings on new items. They are thus a promising way to solve the new item cold-start problem. | |||||||||||
