Recommending a content curator

• US 10,528,572 B2
• Filed: 08/28/2015
• Issued: 01/07/2020
• Est. Priority Date: 08/28/2015
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Sacheti, Arun, Sun, Yanfeng, Yuen, Aaron Chun Win, Govindarajen, Parthasarathy, Wu, Kun, Cho, Soohoon, Pradhan, Malik Mehdi, Michelis, Alexandre, Vaidya, Gautam Vishwas, Hasham, Karim Amin, Vemuluru, Avinash
• Primary Examiner(s): Ortiz Ditren, Belix M
• Application Number:
  US14/839,385
• Publication Number:
  US 20170060872A1
• Time in Patent Office: 1,593 Days
• Field of Search: None
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

The technology described herein provides an efficient mechanism for quickly analyzing huge amounts of media content to find media content (hereafter “content” or “media content”) that is relevant to a user. The technology analyzes features of a curator to classify curators by interest and/or find curators with similar content recommendations. The curator data can be used to make curator recommendations to users based on the user'"'"'s interests. The technology described herein collects curator data from multiple content sites and analyzes the data to identify curators that recommend similar content on different content sites.

Claims

1. A search server comprising:

a processor; and

a computer storage memory having computer-executable instructions for a monitoring application stored thereon which, when executed by the processor, implement a method for recommending a curator, having one or more content collections, to a user, the method comprising;

receiving a search query from a user through a search interface of an online search engine;

determining that a plurality of content on a first content site is responsive to the query, and the content site is separate from the online search engine;

determining the user does not currently follow any curators on the content site;

based on the determination that the user does not currently follow any curators on the content site, calculating a compatibility score for a first curator from the community of curators, the compatibility score measuring a conformity between a style for the user and a style for the first curator, wherein the style for the user is determined using interaction data describing interactions with content on a second content site that is separate from the first content site, and wherein the style of the first curator is determined from content recommendations made by the first curator;

determining the compatibility score satisfies a recommendation requirement; and

upon said determining the compatibility score satisfies the recommendation requirement, outputting for display to the user a recommendation for the first curator.

2. The search server of claim 1, wherein the recommendation is output on a search results page provided in response to the query.

3. The search server of claim 1, wherein the style for the user is determined using a computer learning model that assigns one or more classifications to the user taking relationships the user has with other curators on one or more content sites other than the first content site as input.

4. The search server of claim 1, wherein the method further comprises building a curator co-relation model that comprises a similarity score for curators in different content sites, wherein the similarity score is used to calculate the compatibility score.

5. The search server of claim 4, wherein the similarity score between a first curator in the first content site and a second curator in a second content site is based on an amount of common content being recommended by both the first curator and the second curator.

6. The search server of claim 4, wherein the similarity score between a first curator in the first content site and a second curator in a second content site is based on an amount of unique users that have relationships with both the first curator and the second curator.

7. The search server of claim 1, wherein the recommendation for the first curator is output through a personal digital assistant interface after the user navigates to the first content site in response to clicking on a search result.

8. A method for recommending a curator to a user comprising:

determining that a user has navigated to a first content site that is associated with a community of curators;

determining that the user has not previously visited the first content site;

based on the determination that the user has not visited the first content site previously, calculating a compatibility score for a first curator from the community of curators, the compatibility score measuring a conformity between a style for the user and a style for the first curator, wherein the style for the user is determined using interaction data describing interactions with content on a second content site that is separate from the first content site and relationships with content curators on the second site, and wherein the style of the first curator is determined from content recommendations made by the first curator;

determining the compatibility score satisfies a recommendation requirement; and

outputting for display to the user a recommendation for the first curator.

9. The method of claim 8, wherein the compatibility score comprises a similarity score and a curator classification.

10. The method of claim 9, wherein the curator classification is generated by a classifier application that analyzes curator features to assign one or more curator classifications to the curator.

11. A method of claim 10, wherein one of the curator features is derived from an automated analysis of images of content within a collection generated by the curator.

12. The method of claim 8, wherein the recommendation for the curator is output through an interface generated by a personal assistant application.

13. The method of claim 8, wherein the method further comprises communicating the recommendation for the curator to the content site, and wherein the recommendation for the curator is output through an interface generated by the content site.

14. A method for recommending a curator to a user, the method comprising:

determining that a user has relationships with a plurality of curators from multiple content sites;

determining that a user has not visited a first content site previously;

based on the determination that a user has not visited the first content site, determining based on features of the plurality of curators that a new curator from the first content site has a similarity score with the plurality of curators that is above a threshold score, wherein the user does not currently have a relationship with the new curator; and

outputting for display a recommendation for the new curator.

15. The method of claim 14, wherein the similarity score is calculated using a vector analysis where features of individual curators form vector dimensions.

16. The method of claim 14, wherein the similarity score is calculated using users-in-common as an input, wherein users-in-common is an amount of users two curators have in common.

17. The method of claim 14, wherein the recommendation is output by a personal assistant application in a context where the user is seeking content related to the curator.

18. The method of claim 14, wherein a relationship between the user and a curator is formed when the user affirmatively follows the curator.