本文属于CF based中的MF系列. 本质上是利用了多个矩阵转换构建了最后的estimated score. 其中最主要的motivation是: 对每个上下文因子中的每个取值都学习出一个contextual latent factor vector (clfv for short). 而最终的上下文隐因子向量(只有一个)是通过加权这些单个的上下文因子的clfv生成的.(这样做的好处是,对于不同用户,即使是相同的上下文组合,他们的clfv也可能不一样)
其他文章学习的contextual latent factor vector一般是针对的是上下文因子的组合(一个组合代表一个真实的上下文), 如:考虑两个上下文因子,每个因子有3种取值,那么一共就有3X3=9种上下文,对应的就要学习出9各contextual latent factor vectors, 至于每个vector的维度,it is tuned empirically.
仔细分析上面两种方法就可以得出文中所谓的, traditional methods 的关于上下文的隐因子向量对于所有人都是一样的(即,只要上下文一样,那么clfv就一样),是因为上下文组合个数跟学习的clfv个数是一样的;而本文就是在这一点上做的改进, 本文的clfv是要多一个单个隐因子向量线性组合步骤才能得出的,每个人的权值向量不一样,线性组合后的clfv自然也不一样.
图示:
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| COT: Contextual Operating Tensor for Context-aware Recommender Systems | |||||||||||
| 文献作者 | Qiang Liu, Shu Wu, Liang Wang | ||||||||||
| 文献发表年限 | 2015 | ||||||||||
| 文献关键字 | |||||||||||
| MF; Tensor; nlp | |||||||||||
| 摘要描述 | |||||||||||
| With rapid growth of information on the Internet, recommender systems become fundamental for helping users alleviate the problem of information overload. Since contextual information can be used as a significant factor in modeling user behavior, various context-aware recommendation methods are proposed. However, the state-of-the-art context modeling methods treat contexts as other dimensions similar to the dimensions of users and items, and cannot capture the special semantic operation of contexts. On the other hand, some works on multi-domain relation prediction can be used for the context-aware recommendation, but they have problems in generating recommendation under a large amount of contextual information. In this work, we propose Contextual Operating Tensor (COT) model, which represents the common semantic effects of contexts as a contextual operating tensor and represents a context as a latent vector. Then, to model the semantic operation of a context combination, we generate contextual operating matrix from the contextual operating tensor and latent vectors of contexts. Thus latent vectors of users and items can be operated by the contextual operating matrices. Experimental results show that the proposed COT model yields significant improvements over the competitive compared methods on three typical dataset- s, i.e., Food, Adom and Movielens-1M datasets. | |||||||||||
