Microsoft Technology Licensing, LLC

Madhulekha Arunmozhi

Manas Haribhai Somaiya

Ian Ackerman

Sarah Yan Xing

Birjodh S Tiwana

Vivek Yogesh Tripathi

Ali Sadiq Mohamed

Xin Hu

Shunlin Liang

Manish Swaminathan

Prateek Sachdev

Strahinja Markovic

Muhammad Hassan Khan

Caitlin Marie O'Connor

Zeesha Currimbhoy

Prashant Saxena

Described herein is a technique for surfacing content for users of a connection network application. The technique involves performing a first query to fetch recently-impressed items viewed by a user of the connection network application and, concurrently with the first query, performing a second query of search nodes to generate a set of search node results. The technique allows for filtering the recently-impressed items from the set of search node results to generate a candidate set and applying a freshness factor to the candidate set. Updates are provided to a user display, based on results of applying the freshness factor to the candidate set.

Impression discounting for followfeed

Approximating a more complex multi-objective feed item scoring model using a less complex single objective feed item scoring model in a multistage feed ranking system of an online service. The disclosed techniques can facilitate multi-objective optimization for personalizing and ranking feeds including balancing personalizing a feed for viewer experience, downstream professional or social network effects, and upstream effects on content creators. The techniques can approximate the multi-objective model—that uses a rich set of machine learning features for scoring feed items at a second pass ranker in the ranking system—with the more lightweight, single objective model—that uses fewer machine learning features at a first pass ranker in the ranking system. The single objective model can more efficiently score a large set of feed items while maintaining much of the multi-objective model's richness and complexity and with high recall at the second pass ranking stage.

Multistage feed ranking system with methodology providing scalable multi-objective model approximation

Approximating a more complex multi-objective feed item scoring model using a less complex single objective feed item scoring model in a multistage feed ranking system of an online service. The disclosed techniques can facilitate multi-objective optimization for personalizing and ranking feeds including balancing personalizing a feed for viewer experience, downstream professional or social network effects, and upstream effects on content creators. The techniques can approximate the multi-objective model&#x2014;that uses a rich set of machine learning features for scoring feed items at a second pass ranker in the ranking system&#x2014;with the more lightweight, single objective model&#x2014;that uses fewer machine learning features at a first pass ranker in the ranking system. The single objective model can more efficiently score a large set of feed items while maintaining much of the multi-objective model's richness and complexity and with high recall at the second pass ranking stage.

Methods, systems, and computer programs are presented for reducing latency for providing a user feed containing one or more posts. One method includes operations for receiving a request to access the user feed and for performing a first query to search posts. The first query uses a first time horizon delimiting a creation time of posts and a first maximum number of posts selected for ranking. The posts from the first query are sent to the client device for presentation on a user interface. Further, a second query is performed to search posts, where the second query uses a second time horizon that is greater than the first time horizon and a second maximum number of posts for ranking that is greater than the first maximum number of posts. The posts from the first query and the second query are merged and sent to the client device for presentation.

Latency optimization for online service feed

A feature importance score for a target machine learning feature of a target machine learning model used in a multistage feed ranking system for scoring feed items is supplemented with a feature computing resource cost. The feature computing resource cost represents the cost of using the target feature in the target model in terms of computing resources such as CPU, memory, network resources, etc. A tradeoff between feature importance and feature computing resource cost can be made to decide whether to have the target machine learning model use or not use the target machine learning feature in production, thereby improving the production multistage feed item ranking system and solving the technical problem of determining which machine learning features of a machine learning model represent the best tradeoff between feature importance and feature computing resource cost.

Multistage feed ranking system with methodology providing scoring model optimization for scaling

Growing community of inventors

Sunnyvale, CA, United States of America

Madhulekha Arunmozhi