所谓的view data指的是用户click 行为，区别于购买行为。
思想很简单也很老：用户对purchase item的preference要高于clicked item, 同时 clicked itemdpreference要高于unclick item.
贡献可能在于：结合了eALS算法.......
Loss function由三个部分组成：R_ui, R_ui > V_ui, V_ui > others; 这里的pairwise值得借鉴：(a-(R_ui - V_ui))^2： 有点Metric Learning的意思
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| Improving Implicit Recommender Systems with View Data | |||||||||||
| 文献作者 | Jingtao Ding | ||||||||||
| 文献发表年限 | 2018 | ||||||||||
| 文献关键字 | |||||||||||
| 摘要描述 | |||||||||||
| Most existing recommender systems leverage theprimary feedback data only, such as the purchase records in E-commerce. In this work, we additionally integrate view data into implicit feedback based recommender systems (dubbed asImplicit Recommender Systems). We propose to model the pairwise ranking relations among purchased, viewed, and non-viewed interactions, being more effective and flexible than typical pointwise matrix factorization (MF) methods. However, such a pairwise formulation poses efficiency challenges in learning the model. To address this problem, we design a new learning algorithm based on the element-wise Alternating Least Squares(eALS) learner. Notably, our algorithm can efficiently learn model parameters from the whole user-item matrix (including all missing data), with a rather low time complexity that is dependent on the observed data only. Extensive experiments on two real-worlddatasets demonstrate that our method outperforms several state-of-the-art MF methods by10%28:4%. Our implementation is available at: https://github.com/dingjingtao/ViewenhancedALS. | |||||||||||
