US 11,308,096 B2Granted

Bias quotient measurement and debiasing for recommendation engines

AssigneeRovi Guides, Inc.

InventorsRajendran Pichaimurthy, Madhusudhan Srinivasan

Life of this patent

Estimating the term…

Loading patent…

Abstract

Systems and methods for debiasing a recommendation engine are disclosed herein. A search query associated with a user profile is received at a recommendation engine. Control circuitry generates a result set of items of content based on the search query and generates a bias score for a content attribute based on the result set. The control circuitry also generates a time-averaged bias score for the content attribute based on a plurality of search queries associated with the user profile. Based on the bias score and the time-averaged bias score, the control circuitry determines whether a bias is signaled for the content attribute. Finally, the control circuitry outputs, for display via a computing device, the result set or a debiased result set based on a result of the determination of whether the bias is signaled.

Claims20

Dependent claims

A method for debiasing a recommendation engine, comprising:

receiving, at a recommendation engine, a search query associated with a user profile;
generating a result set of items of content based on the search query;
generating, based on the result set, a bias score for a content attribute;
receiving, at the recommendation engine, a plurality of search queries associated with the user profile;
for each of the plurality of search queries associated with the user profile:
(i) generating a respective result set of items of content;
and
(ii) generating a respective bias score based on the respective result set;
generating a time-averaged bias score for the content attribute based on computing an average of all the respective bias scores generated for the plurality of search queries over a period of time;
determining, based on the bias score and the time-averaged bias score, whether a bias is signaled for the content attribute;
and
outputting, for display via a computing device, the result set or a debiased result set based on a result of the determination of whether the bias is signaled.

9 more elements

A system for debiasing a recommendation engine, comprising:

a communication port configured to receive, at a recommendation engine, a search query associated with a user profile and a plurality of search queries associated with the user profile;
memory configured to store instructions;
and
processing circuitry configured to execute the instructions to:
generate a result set of items of content based on the search query;
generate, based on the result set, a bias score for a content attribute;
for each of the plurality of search queries associated with the user profile:
(i) generate a respective result set of items of content;
and
(ii) generate a respective bias score based on the respective result set;
generate a time-averaged bias score for the content attribute based on computing an average of all the respective bias scores generated for the plurality of search queries over a period of time;
determine, based on the bias score and the time-averaged bias score, whether a bias is signaled for the content attribute;
and
output, for display via a computing device, the result set or a debiased result set based on a result of the determination of whether the bias is signaled.

11 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/370,181Filed 29 Mar 2019
Published: US 2020/0311090 A11 Oct 2020
Granted: 19 Apr 2022
Term ends: Not estimatedThe source records don’t support an estimate.

Classification4 codes

G06F 16/24578
With adaptation to user needs: using ranking
G06F 16/248
Querying: presentation of query results
G06F 16/9535
Querying, e.g. by the use of web search engines: search customisation based on user profiles and personalisation
G06F
Computing and data processing

Family

US 2020/0311090 A1Pre-grant publication · published 1 Oct 2020
US 11,308,096 B2This 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