Introduction 结构
1. 个性化推荐很流行(prevalent), 有优势; MF是基础; 早期MF关注explicit feedback; 于是他们属于rating prediction一类; 这类方法的优势
2. 显示反馈are not always avaliable; 隐式反馈优势以及其挑战(偏向正样本)
3. 为解决负样本缺失问题; 方法一: all missing as negative,其缺点是学习效率,不适合在线性运行; 也有人考虑过提高MF based on implicit feedback的效率,但它不是最优的,也不切合实际,具体说来...它假设missing服从高斯分布, 这种假设是不符合实际情况的
4. How do we do, 主要是要解决基于隐式反馈推荐系统训练的有效性和高效性: 这篇文章提出的方法不是高斯分布的,反而是基于物品流行度的.
总结: 这是一篇关注学习效率的文章. 从负样本角度出发 : sample based learning and whole-data based learning
本文核心思想(model学习技巧)
在加权矩阵分解的基础上, 将正样本的的权重和unknown样本的权重分离开来, 然后在unknown 样本空间实施一些技巧,从而加速学习. 需要注意的是,他是把这样的分离思想引用到了Generic Element-wise ALS 上.
本文之所以没有学习效率(步长)参数的原因,是它是ALS-based 区别于GD(SGD和BGD)
本文认为(同 mproving pairwise learning for item recommendation from implicit feedback. In WSDM 2014 ), 越流行的物品,用户越没有点击,越有可能是true negative sample
在线跟新的技巧:
仅仅更新该新record的用户和被访问的item. 因为它的影响是有限的(这是一个假设)
实验时,假定某一部分数据为new data, 然后融入原来的trained model
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| Fast Matrix Factorization for Online Recommendation with Implicit Feedback | |||||||||||
| 文献作者 | Xiangnan He; Hanwang Zhang | ||||||||||
| 文献发表年限 | 2016 | ||||||||||
| 文献关键字 | |||||||||||
| Matrix Factorization, Implicit Feedback, Item Recommen- dation, Online Learning, ALS, Coordinate Descent | |||||||||||
| 摘要描述 | |||||||||||
| This paper contributes improvements on both the effective- ness and efficiency of Matrix Factorization (MF) methods for implicit feedback. We highlight two critical issues of ex- isting works. First, due to the large space of unobserved feedback, most existing works resort to assign a uniform weight to the missing data to reduce computational com- plexity. However, such a uniform assumption is invalid in real-world settings. Second, most methods are also designed in an offline setting and fail to keep up with the dynamic nature of online data. We address the above two issues in learning MF models from implicit feedback. We first propose to weight the miss- ing data based on item popularity, which is more effective and flexible than the uniform-weight assumption. However, such a non-uniform weighting poses efficiency challenge in learning the model. To address this, we specifically de- sign a new learning algorithm based on the element-wise Alternating Least Squares (eALS) technique, for efficiently optimizing a MF model with variably-weighted missing data. We exploit this efficiency to then seamlessly devise an incre- mental update strategy that instantly refreshes a MF model given new feedback. Through comprehensive experiments on two public datasets in both offline and online protocols, we show that our eALS method consistently outperforms state-of-the-art implicit MF methods. Our implementation is available at https://github.com/hexiangnan/sigir16-eals. | |||||||||||
