	US 12,190,209 B2
	Machine-learned models incorporating sequence encoders that operate on bag of words input
Meng Meng, San Jose, CA (US); Daniel Sairom Krishnan Hewlett, Sunnyvale, CA (US); Sriram Vasudevan, Mountain View, CA (US); and Vitaly Abdrashitov, Sunnyvale, CA (US)
Assigned to Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on May 6, 2021, as Appl. No. 17/313,560.
Prior Publication US 2022/0358398 A1, Nov. 10, 2022
Int. Cl. G06N 20/00 (2019.01); G06F 16/951 (2019.01); G06F 16/955 (2019.01); G06F 18/2113 (2023.01); H04L 67/02 (2022.01)

CPC G06N 20/00 (2019.01) [G06F 16/951 (2019.01); G06F 16/955 (2019.01); G06F 18/2113 (2023.01); H04L 67/02 (2013.01)]

17 Claims

1. A method comprising:

identifying a plurality of tokens that are associated with online activities of an entity;

for each token in the plurality of tokens, identifying a standardized identifier that is uniquely associated with said each token;

wherein each standardized identifier in a plurality of standardized identifiers uniquely identifies one of the tokens in the plurality of tokens;

identifying a plurality of machine-learned embeddings that correspond to the plurality of tokens;

based on one or more ordering criteria that are independent of the temporal occurrence of the online activities of the entity, determining an order of the plurality of machine-learned embeddings;

wherein determining the order is based on the values of the plurality of standardized identifiers;

based on the order, inputting the plurality of machine-learned embeddings to a sequence encoder that generates output;

based on the output, generating, by a machine learned model that includes the sequence encoder, a score;

selecting a content item based on the score;

causing the content item to be transmitted over a computer network to a computing device;

wherein the method is performed by one or more computing devices.