Facebook, Inc.

Lu Zheng

Wei Guang Wu

Tianshi Gao

Shyamsundar Rajaram

Hao Zhang

Feng Yan

David Michael Viner

Jiesi Luo

An online system tracks stores information identifying content provided by third party systems and accessed by online system users as well as interactions with advertisements performed by online system users. When the online system identifies an opportunity to present an advertisement to a viewing user, the online system identifies content from third party systems accessed by the viewing user and content from third party systems accessed by additional online system users who interacted with advertisements. A score is computed for various advertisements based at least in part on correlations between content from third party systems accessed by the viewing user and content from third party systems accessed by additional online system users who interacted with advertisements. The online system selects candidate advertisements to evaluate for presentation to the viewing user based on the scores.

Selecting content for presentation to online system users based on correlations between content accessed by users via third party systems and interactions with online system content

In one embodiment, a method includes, receiving a request, from a client system of a first user, to access a marketplace including products offered for sale by a second user. The method includes, by the computing device, filtering a set of product listings based on multiple respective product-listing embeddings and a content-interaction embedding associated with the first user. Each product listing includes a description of one of the products in the marketplace. The method includes ranking each product listing in the filtered set based at least on a product-score representing a likelihood of the first user interacting with the respective product. The product-score is based on interaction information associated with the first user, product information associated with the product, and sparse information associated with the first user. The method includes sending, to the client system of the first user, a subset of the ranked product listings.

Marketplace feed ranking on online social networks

Growing community of inventors

Palo Alto, CA, United States of America

Lu Zheng