Methods and apparatus for automatically providing personalized search results

Patent metadata

Field	Value
Publication number	US-12019639-B2
Application number	US-202318101188-A
Country	US
Kind code	B2
Filing date	Jan 25, 2023
Priority date	Jul 30, 2019
Publication date	Jun 25, 2024
Grant date	Jun 25, 2024

Abstract

This application relates to apparatus and methods for generating preference profiles that may be used to rank search results. In some examples, a computing device obtains browsing session data and determines items that were engaged, such as items that were viewed or clicked. The computing device obtains item property data, such as product descriptions, for the items, and applies a dependency parser to the item property data to identify portions that include certain words, such as nouns or adjectives, which are then identified as attributes. The computing device generates attribute data identifying portions of the item property data as item attributes. In some examples, the computing device applies one or more machine learning algorithms to the session data and/or search query to identify item attributes. The computing device may generate a profile that includes the item attributes, and may rank search results based on the attribute data, among other uses.

First claim

What is claimed is:

1. A system comprising:
a memory resource, storing instructions; and
one or more processors coupled to the memory resource, the one or more processors being configured to execute the instructions to:
obtain a session data of a user, the session data identifying and
characterizing one or more previous instances the user engaged, via a computing device of the user, with each item of a set of items provided by a retailer;
obtain, from catalogue data, item property data associated with a previously engaged item of the set of items;
determine at least portion of the item property data for the previously engaged item based on one or more words associated with the previously engaged item;
receive, from the computing device of the user via a server, a search request identifying a plurality of search terms for a currently searched item, the currently searched item being engaged with by the user subsequent to the previously engaged item;
generate a plurality of word embeddings and a plurality of character embeddings for the session data of the user, the catalogue data, and the plurality of search terms of the search request by applying a first neural network;
apply a second neural network to combine each of the plurality of word embeddings to the each of the corresponding plurality of character embeddings;
apply a third neural network to the output of the second neural network and in response to applying the third neural network, generate first attribute data identifying an attribute of the previously engaged item based on the determined at least portion of the item property data and associate the first attribute data with the user;
and based at least on the first attribute data and the search request, generate search results.

2. The system of claim 1 , wherein the computing device is configured to:
based at least on the search request, implement a set of operations that generate second attribute data, the set of operations including:
determining at least portion of the plurality of search terms based on applying a dependency parser to the plurality of search terms;
generating second attribute data identifying at least a second attribute based on the determined at least portion of the plurality of search terms; and
storing the first attribute data and the second attribute data in a database.

3. The system of claim 2 , wherein determining the at least portion of the plurality of search terms comprises determining at least one noun and at least one adjective of the plurality of search terms.

4. The system of claim 2 , wherein generating the search results for the search request includes:
ranking the search results based on at least one of the first attribute data and the second attribute data; and
transmitting the ranked search results to the server.

5. The system of claim 2 , wherein the dependency parser is a part-of-speech tagger.

6. The system of claim 1 , wherein the at least portion of the item property data comprises an item description for the previously engaged item.

7. The system of claim 1 , wherein the session data identifies that the previously engaged item has been viewed.

8. The system of claim 1 , wherein determining the at least portion of the item property data for the previously engaged item comprises determining at least one item type.

9. A method comprising:
obtaining a session data of a user, the session data identifying and characterizing one or more previous instances the user engaged, via a computing device of the user, with each item of a set of items provided by a retailer;
obtaining, from catalogue data, item property data associated with a previously engaged item of the set of items;
determining at least portion of the item property data for the previously engaged item based on one or more words associated with the previously engaged item;
receiving, from the computing device of the user via a server, a search request identifying a plurality of search terms for a currently searched item, the currently searched item being engaged with by the user subsequent to the previously engaged item;
generating a plurality of word embeddings and a plurality of character embeddings for the session data of the user, the catalogue data, and the plurality of search terms of the search request by applying a first neural network;
applying a second neural network to combine each of the plurality of word embeddings to the each of the corresponding plurality of character embeddings;
applying a third neural network to the output of the second neural network and in response to applying the third neural network, generate first attribute data identifying an attribute of the previously engaged item based on the determined at least portion of the item property data and associate the first attribute data with the user; and
based at least on the first attribute data and the search request, generating search results.

10. The method of claim 9 , further comprising:
based at least on the search request, implementing a set of operations that generate second attribute data, the set of operations including:
determining at least portion of the plurality of search terms based on applying a dependency parser to the plurality of search terms;
generating second attribute data identifying at least a second attribute based on the determined at least portion of the plurality of search terms; and
storing the first attribute data and the second attribute data in a database.

11. The method of claim 10 , wherein determining the at least portion of the plurality of search terms comprises determining at least one noun and at least one adjective of the plurality of search terms.

12. The method of claim 10 , wherein generating the search results for the search request includes:
ranking the search results based on at least one of the first attribute data and the second attribute data; and
transmitting the ranked search results to the server.

13. The method of claim 9 , wherein the at least portion of the item property data comprises an item description for the previously engaged item.

14. The method of claim 9 , wherein determining the at least portion of the item property data for the previously engaged item comprises determining at least one item type.

15. A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by at least one processor, cause a device to perform operations comprising:
obtaining a session data of a user, the session data identifying and characterizing one or more previous instances the user engaged, via a computing device of the user, with each item of a set of items provided by a retailer;
obtaining, from catalogue data, item property data associated with a previously engaged item of the set of items; determining at least portion of the item property data for the previously engaged item based on one or more words associated with the previously engaged item;
receiving, from the computing device of the user via a server, a search request identifying a plurality of search terms for a currently searched item, the currently searched item being engaged with by the user subsequent to the previously engaged item;
generating a plurality of word embeddings and a plurality of character embeddings for the session data of the user, the catalogue data, and the plurality of search terms of the search request by applying a first neural network;
applying a second neural network to combine each of the plurality of word embeddings to the each of the corresponding plurality of character embeddings;
applying a third neural n

Assignees

• Walmart Apollo Llc

Inventors

• Iyer Rahul
• Wadhwa Soumya
• Guo Stephen Dean
• Achan Kannan

Classifications

• G06F16/22

  Indexing; Data structures therefor; Storage structures · CPC title

• G06F16/9535

  Search customisation based on user profiles and personalisation · CPC title

• G06F16/24578 Primary

  using ranking · CPC title