Automatically generating baskets of items to be recommended to users of an online system

Patent metadata

Field	Value
Publication number	US-12591918-B2
Application number	US-202318202768-A
Country	US
Kind code	B2
Filing date	May 26, 2023
Priority date	May 26, 2023
Publication date	Mar 31, 2026
Grant date	Mar 31, 2026

Abstract

Embodiments relate to automatically generating a basket of items to be recommended to a user of an online system. The online system communicates a basket opportunity to a group of retailers, wherein the basket opportunity defines a plurality of item categories each associated with a respective item to be included in a basket. The online system receives, from each retailer in response to the basket opportunity, a respective bid of a plurality of bids for the basket opportunity. The online system applies a computer model to each bid to determine a score for each bid and selects a winning bid for the user based on determined scores for the bids. For each item category, the online system populates the basket with a respective item from a catalog of a retailer that is associated with the winning bid. The online system then presents the basket with items to the user.

First claim

What is claimed is:

1 . A method comprising, at a computer system comprising a processor and a computer-readable medium:
communicating, via a network and to a group of devices associated with a group of entities, a signal indicative of a conversion opportunity, wherein the conversion opportunity defines a plurality of item categories each associated with a respective item to be included in a set of items associated with the plurality of item categories that share a common theme; responsive to the communicated signal, receiving, via the network and from each device associated with a respective entity in the group of entities, a respective offer of a plurality of offers for the conversion opportunity; applying a machine-learning model to each of the plurality of offers and information about one or more items of each of the plurality of item categories that are converted by a user of an online system over a defined time period to generate a respective score of a plurality of scores for each of the plurality of offers, wherein the machine-learning model is trained to predict a likelihood of the user selecting the set of items provided by the respective entity if presented given information about the plurality of item categories, the respective score indicative of the predicted likelihood; selecting, using the plurality of scores, an offer of the plurality of offers; populating the set of items, for each of the plurality of item categories, with a respective item of a plurality of items from a database of an entity from the group of entities that is associated with the selected offer; presenting the set of items to the user, wherein presenting the set of items to the user causes a device associated with the user to display a user interface with information about the set of items, an icon representing the set of items, and an option to add the set of items displayed at the user interface into a cart of the user; receiving, via the network and from the device associated with the user, information about adding the set of items into the cart and placing an order including the set of items; responsive to placing the order, assigning a servicing of the order to a picker that is a fully-autonomous robot; upon assigning the servicing of the order, instructing, via collection instructions stored at the computer-readable medium and executed by the processor, the picker operating as the fully-autonomous robot to collect the set of items in a retailer location; physically collecting, by the picker operating as the fully-autonomous robot and using the collection instructions, the set of items in the retailer location; upon collecting the set of items in the retailer location, controlling, via navigation instructions stored at the computer-readable medium and executed by the processor, a movement of the fully-autonomous robot along with an autonomous vehicle from the retailer location to a delivery location associated with the user; and moving, along a navigation route identified using the navigation instructions, the fully-autonomous robot along with the autonomous vehicle from the retailer location to the delivery location for delivering the set of items to the user at the delivery location.

2 . The method of claim 1 , further comprising:
feeding a collection of items purchased by one or more users of the online system into a large language model (LLM) to generate the plurality of item categories for the conversion opportunity.

3 . The method of claim 1 , further comprising:
requesting a large language model (LLM) to provide a textual description of an intent of the conversion opportunity based on a prompt input into the LLM, wherein the prompt includes information about the plurality of item categories; and communicating the textual description of the intent of the conversion opportunity to the group of entities.

4 . The method of claim 3 , further comprising:
determining the group of entities by filtering a collection of items from a plurality of entities associated with the online system using at least one of information about the plurality of item categories and the textual description of the intent of the conversion opportunity.

5 . The method of claim 1 , further comprising:
generating, by the machine-learning model, the respective score for each of the plurality of offers based on the predicted likelihood of the user selecting the set of items provided by the respective entity of the group of entities that is associated with each of the plurality of offers.

6 . The method of claim 1 , further comprising:
identifying, by the machine-learning model and for the user, the respective item from the database for each of the plurality of item categories.

7 . The method of claim 1 , further comprising:
identifying, by the machine-learning model and for the user, the respective item from the database for each of the plurality of item categories, based on one or more attributes of that item category and information about a set of items of that item category purchased by the user over a defined time period.

8 . The method of claim 1 , further comprising:
providing, by the machine-learning model, a score to each candidate item of a plurality of candidate items from the database of the entity for each of the plurality of item categories, based on a predicted likelihood of the user selecting a set of items populated with that candidate item; and identifying, by the machine-learning model, the respective item for populating the set of items based on the score of each candidate item.

9 . The method of claim 1 , further comprising:
applying, by the machine-learning model, a multi-arm bandit algorithm to a plurality of candidate items from the database of the entity for each of the plurality of item categories to select the respective item for populating the set of items.

10 . The method of claim 1 , further comprising:
identifying, by the machine-learning model, a profile of the user based at least in part on information about a collection of items purchased by the user over a defined time period; and prompting a large language model (LLM) with the profile of the user and information about the populated set of items to generate a description of the set of items for presentation to the user at the user interface.

11 . The method of claim 1 , wherein each entity in the group of entities provides the respective offer depending on an objective of an advertising campaign of the online system.

12 . A computer program product comprising a non-transitory computer-readable storage medium having instructions encoded thereon that, when executed by a processor, cause the processor to perform steps comprising:
communicating, via a network and to a group of devices associated with a group of entities, a signal indicative of a conversion opportunity, wherein the conversion opportunity defines a plurality of item categories each associated with a respective item to be included in a set of items associated with the plurality of item categories that share a common theme; responsive to the communicated signal, receiving, via the network and from each device associated with a respective entity in the group of entities, a respective offer of a plurality of offers for the conversion opportunity; applying a machine-learning model to each of the plurality of offers and information about one or more items of each of the plurality of item categories that are converted by a user of an online system over a defined time period to generate a respective score of a plurality of scores for each of the plurality of offers, wherein the machine-learning model is trained to predict a lik

Assignees

• Maplebear Inc

Inventors

• Archak Shrikar
• Prasad Shishir Kumar

Classifications

• G06Q30/08

  Auctions · CPC title

• G06Q30/0631 Primary

  Recommending goods or services · CPC title