Generating user recommendations

Patent metadata

Field	Value
Publication number	US-9665663-B2
Application number	US-201414573756-A
Country	US
Kind code	B2
Filing date	Dec 17, 2014
Priority date	Dec 17, 2014
Publication date	May 30, 2017
Grant date	May 30, 2017

Abstract

Briefly, embodiments of methods and/or systems of providing relevant and diverse recommendations are disclosed. For one embodiment, as an example, a system may extract structured and/or semi-structured parameters from web resources obtained from interaction logs comprising records of browsing sessions. Content from extracted parameters may be compared, using an ontology, to find relationships among web resources and query resources.

First claim

What is claimed is:

1. A method comprising:
generating cross-domain recommendations for presentation to one or more users based at least in part on real-time web browsing, on a list of Web sites or Web pages visited by the one or more users, and on structured and/or semi-structured parameters stored in memory and extracted from one or more Web sites or Web pages visited while browsing, the generating comprising identifying semantic bridges for use in generating the cross-domain recommendations, wherein
identifying semantic bridges comprises populating an ontology based, at least in part, on the list of Web sites or Web pages visited by the one or more users and on the structured and/or semi-structured parameters extracted from the one or more Web sites or Web pages visited while browsing.

2. The method of claim 1 , wherein the generating comprises creating feature signal vectors for use in machine learning, wherein the feature signal vectors are based at least in part on the list of Web sites or Web pages visited by the one or more users and on the structured and/or semi-structured parameters extracted from the one or more Web sites or Web pages visited while browsing.

3. The method of claim 2 , wherein the feature signal vectors comprise one or more classes of non-relevant and/or relevant resources.

4. The method of claim 2 , wherein the feature signal vectors are based at least in part on similarity among semantic types.

5. The method of claim 1 , wherein the recommendations for presentation to one or more users are based at least in part on results of machine learning generated from the real-time web browsing.

6. The method of claim 5 , wherein the machine learning comprises SVM.

7. An apparatus, comprising:
one or more processors to:
generate cross-domain recommendations for display based at least in part on real-time web browsing of a user, on a list of Web sites or Web pages to be visited by the user, and on structured and/or semi-structured parameters to be stored in memory and to be extracted from one or more Web sites or Web pages visited by the user, the generated cross-domain recommendations to comprise identification of semantic bridges via populating an ontology based, at least in part, on the list of Web sites or Web pages visited by the user and on the structured and/or semi-structured parameters to be extracted from the one or more Web sites or Web pages visited by the user.

8. The apparatus of claim 7 , wherein the one or more processors are further to generate feature signal vectors to be used in connection with a machine learning process, wherein the feature signal vectors are to be based at least in part on the list of Web sites or Web pages visited by the user and on the structured and/or semi-structured parameters to be extracted.

9. The apparatus of claim 8 , wherein the feature signal vectors are to be based at least in part on similarity among semantic types.

10. The apparatus of claim 7 , wherein the recommendations are to be based at least in part on machine-learned results to be generated from real-time web browsing.

11. An apparatus comprising:
means for generating cross-domain recommendations for presentation to one or more users based at least in part on real-time web browsing, on a list of Web sites or Web pages visited by the one or more users, and on structured and/or semi-structured parameters stored in memory and extracted from one or more Web sites or Web pages visited while browsing, the means for generating cross-domain recommendations comprising means for identifying semantic bridges, the means for generating cross-domain recommendations to comprise means for populating an ontology, based, at least in part, on at least the list of Web sites or Web pages visited by the one or more users and on the structured and/or semi-structured parameters extracted from the one or more Web sites or Web pages visited while browsing.

12. The apparatus of claim 11 , wherein the means for generating cross-domain recommendations for presentation comprises means for determining feature signal vectors for use in machine learning, the feature signal vectors to be based at least in part on the the list of Web sites or Web pages visited by the one or more users and on the structured and/or semi-structured parameters extracted from the one or more Web sites or Web pages visited while browsing.

13. The apparatus of claim 11 , wherein the means for generating cross-domain recommendations for presentation further comprises means for determining similarity among semantic types of the structured and/or semi structured parameters extracted from the one or more Web sites or Web pages visited while browsing.

14. The apparatus of claim 11 , further comprising means for recommending for presentation to one or more users based at least in part on machine-learned results generated from the real-time web browsing.

Assignees

• Yahoo Inc

Inventors

• Mika Peter
• Blanco Roi
• Hoxha Julia

Classifications

• G06F17/3097 Primary

  Physics · mapped topic

• G06F17/30867

  Physics · mapped topic

• G06N99/005

  Physics · mapped topic

• G06F17/30734

  Physics · mapped topic

• G06N20/10

  using kernel methods, e.g. support vector machines [SVM] · CPC title