SearchSage:在Pinterest学习搜索查询表征

Nikil Pancha | Software Engineer; Andrew Zhai | Software Engineer; Chuck Rosenberg | Head of Advanced Technologies Group; and Jure Leskovec | Chief Scientist, Advanced Technologies Group

Nikil Pancha | 软件工程师; Andrew Zhai | 软件工程师; Chuck Rosenberg | 先进技术组负责人; Jure Leskovec | 先进技术组首席科学家

Pinterest surfaces billions of ideas to people every day, and the neural modeling of embeddings for content, users, and search queries are key in the constant improvement of these machine learning-powered recommendations. Good embeddings — representations of discrete entities as vectors of numbers — enable fast candidate generation and are strong signals to models that classify, retrieve and rank relevant content.

Pinterest每天向人们展示数十亿个想法，而内容、用户和搜索查询的嵌入神经建模是不断改进这些由机器学习驱动的建议的关键。良好的嵌入--离散实体的数字向量表示--能够快速生成候选人，并且是对相关内容进行分类、检索和排名的模型的强烈信号。

We began our representation learning workstream with Visual Embeddings, a convolutional neural network (CNN) based Image representation, then moved toward PinSage, a graph-based multi-modal Pin representation. We expanded into more use cases such as PinnerSage, a user representation based on clustering a user’s past Pin actions, and have since worked with even more entities including search queries, Idea Pins, shopping items and content creators.

我们从Visual Embeddings开始我们的表示学习工作流，这是一个基于卷积神经网络（CNN）的图像表示，然后转向PinSage，一个基于图形的多模式Pin表示。我们扩展到更多的用例，如PinnerSage，一个基于用户过去的Pin行为聚类的用户表征，并在此后与更多的实体合作，包括搜索查询、Idea Pins、购物项目和内容创建者。

In this blog post we focus on SearchSage, our search query representation, and detail how we built and launched SearchSage for search retrieval and ranking to increase relevance of recommendations and engagement in search across organic Pins, Product Pins, and ads. Now used for 15+ use cases, this embedding is one of the most important features in both our organic and ads relevance models, and has led to metric wins such as an 11% increase in 35s+ click-throughs on product Pins in search, and a 42% increase in related searches.

在这篇博文中，我们专注于SearchSage，我们的搜索查询代表，并详细介绍了我们如何建立和推出SearchSage，用于搜索检索和排名，以...