Identifying similar items based on interaction history

• US 10,311,386 B2
• Filed: 07/08/2011
• Issued: 06/04/2019
• Est. Priority Date: 07/08/2011
• Status: Active Grant

• Current Assignee: Netflix, Inc.
• Original Assignee: Netflix, Inc.
• Inventors: Gomez Uribe, Carlos Alberto, Lucas, Eric, Krishnamurthy, Satish Kumar, Pouliot, Christopher Francis
• Primary Examiner(s): Anderson, Folashade
• Application Number:
  US13/179,392
• Publication Number:
  US 20130013372A1
• Time in Patent Office: 2,888 Days
• Field of Search: 705 729
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

One embodiment of the invention sets forth a mechanism for identifying similar items within a plurality of items available to a group of users. The mechanism includes identifying a first count that indicates a number of times any user has interacted with a first item included in the plurality of items and subsequently with any other item, identifying a second count that indicates a number of times any user within the has interacted with a second item included in the plurality of items after interacting with the first item, computing a probability that indicates the likelihood of a particular user interacting with the second item after interacting with the first item based on the first count and the second count, and computing a similarity score that indicates the similarity between the first item and the second item based on the probability and a popularity score associated with the second item.

Claims

1. A computer-implemented method for identifying two or more similar items within a plurality of digital content items available to a group of users, the method comprising:

providing, via a web server, a web-based interface to a group of users for interacting with a plurality of digital content items, wherein the web-based interface is provided to each user of the group of users through a networked device and enables each user to individually interact with each of the plurality of digital content items;

determining, at the web server, a first count that indicates a number of times users within the group of users have interacted with any other item in the plurality of digital content items, after interacting with a first item included in the plurality of digital content items, the group of users interacting with the first item via web content browsers interacting with the web server through a network;

determining, at the web server, a second count that indicates a number of times users within the group of users have interacted with a second item included in the plurality of digital content items, after interacting with the first item, the group of users interacting with the second item via web content browsers interacting with the web server through a network;

computing, by operation of one or more computer processors, a likelihood of a particular user within the group of users interacting with the second item after interacting with the first item, based on the first count and the second count;

computing a similarity score indicative of a similarity between the first item and the second item, based on the computed likelihood and a popularity score associated with the second item; and

transmitting a recommendation for the second item, via a web content browser, to a user that has interacted with the first digital content item, via the web content browser, based on the similarity score.

2. The method of claim 1, wherein the popularity score associated with the second item indicates the likelihood that the particular user will randomly interact with the second item.

3. The method of claim 2, wherein computing the similarity score further comprises:

computing a ratio of the likelihood to the popularity score associated with the second item and a second popularity score associated with the first item; and

combining the computed ratio with the computed likelihood.

4. The method of claim 1, further comprising applying a noise correction function to the similarity score to account for any inaccurate and/or low-volume data that was captured in the first count or the second count.

5. The method of claim 1, wherein the first count and the second count are identified from data that was captured in periods of time when both the first item and the second item were accessible to the group of users.

6. The method of claim 1, wherein for purposes of computing the likelihood, the second count is weighted based on a metric related to a first user within the group of users who has interacted with the first item.

7. The method of claim 6, wherein the metric indicates a strength of the interaction with the first item, wherein when the interaction is a view of the first item, the strength is a duration of time that the first user viewed the first item.

8. The method of claim 6, wherein the first user has also interacted with the second item, and the metric indicates an amount of time that elapsed between a point in time when the first user interacted with the first item and a point in time when the first user interacted with the second item.

9. The method of claim 1, wherein determining the first count comprises analyzing an interaction history associated with the group of users to determine the number of times any user within the group of users had interacted with the first item.

10. The method of claim 1, wherein the first item and the second item each comprise digital media.

11. A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to identify two or more similar items within a plurality of digital content items available to a group of users, by performing the steps of:

providing, via a web server, a web-based interface to a group of users for interacting with a plurality of digital content items, wherein the web-based interface is provided to each user of the group of users through a networked device and enables each user to individually interact with each of the plurality of digital content items;

determining, at web server, a first count that indicates a number of times users within the group of users have interacted with any other item in the plurality of digital content items, after interacting with a first item included in the plurality of digital content items, the group of users interacting with the first item via web content browsers interacting with the web server through a network;

determining, at the web server, a second count that indicates a number of times users within the group of users have interacted with a second item included in the plurality of digital content items, after interacting with the first item, the group of users interacting with the second item via web content browsers interacting with the web server through a network;

computing, by operation of one or more computer processors, a likelihood of a particular user within the group of users interacting with the second item after interacting with the first item, based on the first count and the second count;

computing a similarity score indicative of a similarity between the first item and the second item, based on the computed likelihood and a popularity score associated with the second item; and

transmitting a recommendation for the second item, via a web content browser, to a user that has interacted with the first digital content item, via the web content browser, based on the similarity score.

12. The non-transitory computer-readable medium of claim 11, wherein the popularity score associated with the second item indicates the likelihood that the particular user will randomly interact with the second item.

13. The non-transitory computer-readable medium of claim 12, wherein computing the similarity score further comprises:

computing a ratio of the likelihood to the popularity score associated with the second item and a second popularity score associated with the first item; and

combining the computed ratio with the computed likelihood.

14. The non-transitory computer-readable medium of claim 11, further comprising applying a noise correction function to the similarity score to account for any inaccurate and/or low-volume data that was captured in the first count or the second count.

15. The non-transitory computer-readable medium of claim 11, wherein the first count and the second count are identified from data that was captured in periods of time when both the first item and the second item were accessible to the group of users.

16. The non-transitory computer-readable medium of claim 11, wherein for purposes of computing the likelihood, the second count is weighted based on a metric related to a first user within the group of users who has interacted with the first item.

17. The non-transitory computer-readable medium of claim 16, wherein the metric indicates a strength of the interaction with the first item, wherein when the interaction is a view of the first item, the strength is a duration of time that the first user viewed the first item.

18. The non-transitory computer-readable medium of claim 16, wherein the first user has also interacted with the second item, and the metric indicates an amount of time that elapsed between a point in time when the first user interacted with the first item and a point in time when the first user interacted with the second item.

19. The non-transitory computer-readable medium of claim 11, wherein determining the first count comprises analyzing an interaction history associated with the group of users to determine the number of times any user within the group of users had interacted with the first item.

20. A computer system, comprising:

a processor; and

a memory containing a program that, when executed by the processor, performs an operation for identifying two or more similar items within a plurality of items available to a group of users, the operation comprising;

providing, via a web server, a web-based interface to a group of users for interacting with a plurality of digital content items, wherein the web-based interface is provided to each user of the group of users through a networked device and enables each user to individually interact with each of the plurality of digital content items;

determining, at the web server, a first count that indicates a number of times users within the group of users have interacted with any other item in the plurality of digital content items, after interacting with a first item included in the plurality of digital content items, the group of users interacting with the first item via web content browsers interacting with the web server through a network,

determining, at the web server, a second count that indicates a number of times users within the group of users have interacted with a second item included in the plurality of digital content items, after interacting with the first item, the group of users interacting with the second item via web content browsers interacting with the web server through a network,

computing, by operation of one or more computer processors, a likelihood of a particular user within the group of users interacting with the second item after interacting with the first item, based on the first count and the second count,

computing a similarity score indicative of a similarity between the first item and the second item, based on the computed likelihood and a popularity score associated with the second item, and

transmitting a recommendation for the second item, via a web content browser, to a user that has interacted with the first digital content item, via the web content browser, based on the similarity score.