arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2005.10110cs.IRcs.LG
收藏

M2GRL:一个面向Web推荐系统的多任务多视图图表示学习框架

M2GRL: A Multi-task Multi-view Graph Representation Learning Framework for Web-scale Recommender Systems

Alibaba Group(阿里巴巴集团);Hong Kong Polytechnic University(香港理工大学)

Menghan Wang, Yujie Lin, Guli Lin, Keping Yang, Xiao-ming Wu

摘要:

将图表示学习与多视图数据(边信息)相结合进行推荐是业界的一种趋势。大多数现有的方法可以归类为\emph{multi-view representation fusion} ;它们首先构建一个图,然后将多视图数据集成到图中每个节点的单个紧凑表示中。然而,这些方法在工程和算法方面都引起了人们的关注:1)多视图数据在工业中是丰富和信息丰富的,并且可能超过单个向量的容量,以及2)由于多视图数据通常来自不同的分布,因此可能引入归纳偏差。在本文中,我们使用\emph{multi-view representation alignment}的方法来解决这个问题。特别是,我们提出了一个多任务多视图图表示学习框架(M2 GRL)学习网络规模推荐系统的节点表示从多视图图。M2 GRL为每个单视图数据构建一个图,从多个图中学习多个单独的表示,并执行对齐以建模跨视图关系。M2 GRL选择多任务学习范式来联合学习视图内表示和跨视图关系。此外,M2 GRL应用同方差不确定性自适应调整训练过程中任务的损失权重。我们在淘宝部署了M2 GRL,并在570亿个示例上对其进行训练。根据离线指标和在线A/B测试,M2 GRL明显优于其他最先进的算法。在淘宝网上对多样性推荐的进一步探索表明了利用\method{}产生的多个表示的有效性,我们认为这是一个有前途的方向,为各种不同关注点的行业推荐任务。

英文摘要:

Combining graph representation learning with multi-view data (side information) for recommendation is a trend in industry. Most existing methods can be categorized as \emph{multi-view representation fusion}; they first build one graph and then integrate multi-view data into a single compact representation for each node in the graph. However, these methods are raising concerns in both engineering and algorithm aspects: 1) multi-view data are abundant and informative in industry and may exceed the capacity of one single vector, and 2) inductive bias may be introduced as multi-view data are often from different distributions. In this paper, we use a \emph{multi-view representation alignment} approach to address this issue. Particularly, we propose a multi-task multi-view graph representation learning framework (M2GRL) to learn node representations from multi-view graphs for web-scale recommender systems. M2GRL constructs one graph for each single-view data, learns multiple separate representations from multiple graphs, and performs alignment to model cross-view relations. M2GRL chooses a multi-task learning paradigm to learn intra-view representations and cross-view relations jointly. Besides, M2GRL applies homoscedastic uncertainty to adaptively tune the loss weights of tasks during training. We deploy M2GRL at Taobao and train it on 57 billion examples. According to offline metrics and online A/B tests, M2GRL significantly outperforms other state-of-the-art algorithms. Further exploration on diversity recommendation in Taobao shows the effectiveness of utilizing multiple representations produced by \method{}, which we argue is a promising direction for various industrial recommendation tasks of different focus.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑