Attribute weighting for media content-based recommendation

Patent metadata

Field	Value
Publication number	US-2017161618-A1
Application number	US-201514962297-A
Country	US
Kind code	A1
Filing date	Dec 8, 2015
Priority date	Dec 8, 2015
Publication date	Jun 8, 2017
Grant date	—

Abstract

Techniques are disclosed for automatically assigning weights to attributes of media content based in part on how many users actually viewed or listened to the content, as well as how many users “liked” or otherwise indicated a preference for the content. The content items can be any type of audio or visual media content, such as songs, videos, or movies, as well as written content, such as books, articles, journals, advertisements, or magazines. A first similarity score is determined based on a similarity between user preferences for content items. A second similarity score is determined based on a similarity between one or more common attributes of the content items. These attributes are assigned ratings that represent the number of users who consumed the corresponding content. Next, weights are assigned to each of the attributes based on the first and second similarity scores using, for example, linear equation regression techniques.

First claim

What is claimed is:

1 . A computer-implemented method for generating content-based recommendations, the method comprising:
determining, by a computer processor, a first similarity score based on a statistical measure of similarity between user preferences for each of a first content item and a second content item; determining, by the computer processor, a second similarity score based on a statistical measure of similarity between a first content attribute and a second content attribute; training a predictive model by assigning, by the computer processor, a weight to the first content attribute based on the first and second similarity scores, and a weight to the second content attribute based on the first and second similarity scores; and generating, by the computer processor and using the predictive model, a content-based recommendation for a content item having both the first content attribute and the second content attribute based on the weights.

2 . The method of claim 1 , wherein the statistical measure of similarity between user preferences is based on historical rating data representing a number of users that indicate a preference for the first content item and a number of users that indicate a preference for the second content item.

3 . The method of claim 2 , further comprising:
calculating, by the computer processor, a number of users that indicate a preference for both the first and the second content items based on the historical rating data; and calculating, by the computer processor, a number of users that indicate a preference for either the first content item or the second content item based on the historical rating data, wherein the first similarity score is determined by dividing the number of users that indicate a preference for both the first and the second content items by the number of users that indicate a preference for either the first content item or the second content item.

4 . The method of claim 1 , wherein assigning the weight comprises generating a set of linear equations based on the first and second similarity scores, and applying a regression function to the set of linear equations to solve for the weight, wherein the weight is a factor in the set of linear equations.

5 . The method of claim 1 , further comprising determining, by the computer processor, a third similarity score representing a statistical measure of similarity between the first and second content items as a function of the weights and the second similarity score.

6 . The method of claim 1 , further comprising determining, by the computer processor, the first similarity score based further on a statistical measure of similarity between each of:
the first content item and a third content item, and the second content item and the third media content item.

7 . The method of claim 1 , wherein each of the first content item and the second content item are at least one of digital audio content, digital video content, and printable content.

8 . In an information processing environment, a system comprising:
a storage; and a computer processor operatively coupled to the storage, the computer processor configured to execute instructions stored in the storage that when executed cause the computer processor to carry out a process comprising:
determining a first similarity score based on a statistical measure of similarity between user preferences for each of a first content item and a second content item;
determining a second similarity score based on a statistical measure of similarity between a first content attribute and a second content attribute; and
training a predictive model by assigning a weight to the first content attribute based on the first and second similarity scores, and a weight to the second content attribute based on the first and second similarity scores.

9 . The system of claim 8 , wherein the statistical measure of similarity between user preferences is based on historical rating data representing a number of users that indicate a preference for the first content item and a number of users that indicate a preference for the second content item.

10 . The system of claim 9 , wherein the process further comprises:
calculating a number of users that indicate a preference for both the first and the second content items based on the historical rating data; and calculating a number of users that indicate a preference for either the first content item or the second content item based on the historical rating data, wherein the first similarity score is determined by dividing the number of users that indicate a preference for both the first and the second content items by the number of users that indicate a preference for either the first content item or the second content item.

11 . The system of claim 8 , wherein assigning the weight comprises generating a set of linear equations based on the first and second similarity scores, and applying a regression function to the set of linear equations to solve for the weight, wherein the weight is a factor in the set of linear equations.

12 . The system of claim 8 , wherein the process further comprises determining a third similarity score representing a statistical measure of similarity between the first and second content items as a function of the weights and the second similarity score.

13 . The system of claim 8 , wherein the process further comprises determining the first similarity score based further on a statistical measure of similarity between each of:
the first content item and a third content item, and the second content item and the third media content item.

14 . The system of claim 8 , wherein each of the first content item and the second content item are at least one of digital audio content, digital video content, and printable content.

15 . A non-transitory computer program product having instructions encoded thereon that when executed by one or more computer processors cause the one or more computer processors to perform a process comprising:
determining a first similarity score based on a statistical measure of similarity between user preferences for each of a first content item and a second content item; determining a second similarity score based on a statistical measure of similarity between a first content attribute and a second content attribute; and training a predictive model by assigning a weight to the first content attribute based on the first and second similarity scores, and a weight to the second content attribute based on the first and second similarity scores.

16 . The non-transitory computer program product of claim 15 , wherein the statistical measure of similarity between user preferences is based on historical rating data representing a number of users that indicate a preference for the first content item and a number of users that indicate a preference for the second content item.

17 . The non-transitory computer program product of claim 16 , wherein the process further comprises:
calculating a number of users that indicate a preference for both the first and the second content items based on the historical rating data; and calculating a number of users that indicate a preference for either the first content item or the second content item based on the historical rating data, wherein the first similarity score is determined by dividing the number of users that indicate a preference for both the first and the second content items by the number of users that indicate a preference for either the first content item or the second content item

Assignees

• Adobe Systems Inc

Inventors

• Swaminathan Viswanathan
• Xu Teng
• Mitra Saayan

Classifications

• G06F16/48 Primary

  Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually · CPC title

• G06F16/40 Primary

  of multimedia data, e.g. slideshows comprising image and additional audio data (retrieval of still image data G06F16/50; retrieval of audio data G06F16/60; retrieval of video data G06F16/70) · CPC title

• G06Q30/02

  Marketing; Price estimation or determination; Fundraising · CPC title

• H04N21/252

  Processing of multiple end-users' preferences to derive collaborative data · CPC title

• G06F16/435

  Filtering based on additional data, e.g. user or group profiles · CPC title