US 2019/0171689 A1Published application

Optimizing item display on graphical user interfaces

AssigneeGoogle Inc.

InventorsAhmed Kachkach, Yi-fan Chen, Karthik Lakshmanan, Sally Goldman, Puneet Chopra, Radhika Malpani, Jeremy Shute, Berna Erol

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

A search computing system receives an interaction history for one or more respective users describing interactions with one or more items and generates a co-interaction matrix, each value in the co-interaction matrix representing a number of common users determined to have had the interaction with an item of a corresponding row and an item of the corresponding column of the co-interaction matrix where the value is located. The search computing system generates an embeddings matrix comprising an item embedding value for each of the one or more items by applying matrix factorization to the co-interaction matrix and determines, in response to a search query of a particular user, a user embedding value for the searching user based on the interaction history for the searching user. The search computing system determines a similarity between each search result and user interaction history by comparing the user embedding value against each of the item embedding values.

Claims20

Dependent claims

A computer-implemented method to optimize item display on graphical user interfaces, comprising:

receiving, by one or more computing devices, a request to render a listing of item entries on a user interface of a user computing device;
analyzing, by the one or more computing devices, first item type interaction histories to determine item embedding values by:
receiving the first item type interaction histories for each of two or more users, the first item type interaction histories describing one or more interactions of a first kind by each of the two or more users with one or more items of a first item type;
generating a co-interaction matrix comprising rows and columns corresponding to each of the one or more items and matrix element values each representing a number of users determined to have the first kind of interaction with a row item of the corresponding row and a column item of the corresponding column within the co-interaction matrix;
generating an embeddings matrix comprising an item embedding value for each of the one or more items;
and
extracting the item embedding value for each of the one or more items from the embeddings matrix;
analyzing, by the one or more computing devices, the first item type interaction histories to generate a user embedding value based on the first item type interaction history for a user associated with the user computing device;
generating, by the one or more computing devices, the listing of item entries on the user interface of the user computing device by:
identifying each of the item entries that are responsive to the request to render the listing of item entries, the item entries comprising one or more corresponding items;
retrieving the item embedding value for each of the one or more item results;
determining similarity values corresponding to a similarity between the determined item embedding value for each of the one or more items and the generated user embedding value;
and
ranking the one or more item results based on the determined similarity values, the ranking of the one or more item results comprising a preferred first item in a first position;
and
rendering, by the one or more computing devices, an optimized display of the one or more item results on the user interface of the user computing device, the optimized display comprising the preferred first item in the first position with each remaining item of the ranked one or more item results in a corresponding ranked non-first position.

13 more elements

A computer program product, comprising:

a non-transitory computer-readable medium having computer-executable program instructions embodied thereon that when executed by computer cause the computer to generate user specific search results to display via graphical user interfaces, the computer-executable program instructions comprising:
computer-executable program instructions to receive a request to render a listing of item entries on a user interface of a user computing device;
computer-executable program instructions to analyze first item type interaction histories to determine item embedding values by:
receiving the first item type interaction histories for each of two or more users, the first item type interaction histories describing one or more interactions of a first kind by each of the two or more users with one or more items of a first item type;
generating a co-interaction matrix comprising rows and columns corresponding to each of the one or more items and matrix element values each representing a number of users determined to have the first kind of interaction with a row item of the corresponding row and a column item of the corresponding column within the co-interaction matrix;
generating an embeddings matrix comprising an item embedding value for each of the one or more items;
and
extracting the item embedding value for each of the one or more items from the embeddings matrix;
computer-executable program instructions to analyze the first item type interaction histories to generate a user embedding value based on the first item type interaction history for a user associated with the user computing device;
computer-executable program instructions to generate the listing of item entries on the user interface of the user computing device by:
identifying each of the item entries that are responsive to the request to render the listing of item entries, the item entries comprising one or more corresponding items;
retrieving the item embedding value for each of the one or more item results;
determining similarity values corresponding to a similarity between the determined item embedding value for each of the one or more items and the generated user embedding value;
and
ranking the one or more item results based on the determined similarity values, the ranking of the one or more item results comprising a preferred first item in a first position;
and
computer-executable program instructions to render an optimized display of the one or more item results on the user interface of the user computing device, the optimized display comprising the preferred first item in the first position with each remaining item of the ranked one or more item results in a corresponding ranked non-first position.

14 more elements

A system to optimize item display on graphical user interfaces, comprising:

a storage device;
and
a processor communicatively coupled to the storage device, wherein the processor executes application code instructions that are stored in the storage device to cause the system to:
receive a request to render a listing of item entries on a user interface of a user computing device;
analyze first item type interaction histories to determine item embedding values by:
receiving the first item type interaction histories for each of two or more users, the first item type interaction histories describing one or more interactions of a first kind by each of the two or more users with one or more items of a first item type;
generating a co-interaction matrix comprising rows and columns corresponding to each of the one or more items and matrix element values each representing a number of users determined to have the first kind of interaction with a row item of the corresponding row and a column item of the corresponding column within the co-interaction matrix;
generating an embeddings matrix comprising an item embedding value for each of the one or more items;
and
extracting the item embedding value for each of the one or more items from the embeddings matrix;
analyze the first item type interaction histories to generate a user embedding value based on the first item type interaction history for a user associated with the user computing device;
generate the listing of item entries on the user interface of the user computing device by:
identifying each of the item entries that are responsive to the request to render the listing of item entries, the item entries comprising one or more corresponding items;
retrieving the item embedding value for each of the one or more item results;
determining similarity values corresponding to a similarity between the determined item embedding value for each of the one or more items and the generated user embedding value;
and
ranking the one or more item results based on the determined similarity values, the ranking of the one or more item results comprising a preferred first item in a first position;
and
render an optimized display of the one or more item results on the user interface of the user computing device, the optimized display comprising the preferred first item in the first position with each remaining item of the ranked one or more item results in a corresponding ranked non-first position

16 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/211,209Filed 5 Dec 2018
Priority: 5 Dec 2017
Published: 6 Jun 2019
Term ends: Not estimatedThe source records don’t support an estimate.

Classification7 codes

G06F 16/9538
Querying, e.g. by the use of web search engines: presentation of query results
G06F 3/0482
Based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like…:…
G06F 16/24578
With adaptation to user needs: using ranking
G06F 16/285
Relational databases: clustering or classification

Family

Application62/594,901Us provisional application · filed 5 Dec 2017
US 2019/0171689 A1This patent

From the US filing history. Foreign family members aren’t listed.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key