User characteristic prediction using images posted in online social networks

Patent metadata

Field	Value
Publication number	US-9489592-B2
Application number	US-201414561895-A
Country	US
Kind code	B2
Filing date	Dec 5, 2014
Priority date	Dec 5, 2014
Publication date	Nov 8, 2016
Grant date	Nov 8, 2016

Abstract

Methods and systems provide electronic instructions to a non-transitory electronic storage hardware device to record images uploaded by a user over a computerized network to a social networking site, and to record categories of network site locations to which the images are uploaded by the user. These methods and systems also provide electronic instructions to a computerized electronic image processor hardware device to analyze features within the images to identify content of each of the images, and to determine the user characteristics based on the categories of network site locations to which the images are uploaded by the user and on the content of the images uploaded by the user. Also, such methods and systems provide electronic instructions to the computerized electronic image processor hardware device to output the user characteristics on a graphic user interface hardware device.

First claim

What is claimed is:

1. A method comprising:
recording images uploaded by a user to a social networking site;
recording categories of social networking boards to which said images are uploaded by said user to identify user image board posting behavior;
analyzing features within said images using automated image feature identification to identify user image posted content of each of said images; and
determining user characteristics of said user based only on said user image board posting behavior and said user image posted content.

2. The method according to claim 1 , said analyzing said features within said images generating vectors describing said user image posted content.

3. The method according to claim 1 , said analyzing said features within said images determining an occurrence count of each type of feature in each said image to produce a vector of each said image.

4. The method according to claim 3 , said analyzing said features within said images determining vectors representing said features,
said method further comprising employing a latent semantic index to discover topics making up a length of each said vector.

5. The method according to claim 1 , said user characteristics comprising at least one of age, gender, preferences, political orientation, and interests.

6. The method according to claim 1 , further comprising identifying said features in said images by employing scale-invariant feature transform (SIFT) to said images.

7. A method comprising:
providing electronic instructions to a non-transitory electronic storage hardware device to record images uploaded by a user over a computerized network to a social networking site;
providing electronic instructions to said non-transitory electronic storage hardware device to record categories of social networking boards to which said images are uploaded by said user to identify user image board posting behavior;
providing electronic instructions to a computerized electronic image processor hardware device to analyze features within said images using automated image feature identification to identify user image posted content of each of said images;
providing electronic instructions to said computerized electronic image processor hardware device to determine user characteristics of said user based only on said user image board posting behavior and said user image posted content; and
providing electronic instructions to said computerized electronic image processor hardware device to output said user characteristics of said user on a graphic user interface hardware device.

8. The method according to claim 7 , said computerized electronic image processor hardware device analyzing said features within said images by generating vectors describing said user image posted content.

9. The method according to claim 7 , said computerized electronic image processor hardware device analyzing said features within said images by determining an occurrence count of each type of feature in each said image to produce a vector of each said image.

10. The method according to claim 9 , said computerized electronic image processor hardware device analyzing said features within said images by determining vectors representing said features,
said method further comprising employing a latent semantic index to discover topics making up a length of each said vector.

11. The method according to claim 7 , said user characteristics comprising at least one of age, gender, preferences, political orientation, and interests.

12. The method according to claim 7 , further comprising said computerized electronic image processor hardware device identifying said features in said images by employing scale-invariant feature transform (SIFT) to said images.

13. An image processor device comprising:
specialized image processing circuits; and
network connections operatively connected to said specialized image processing circuits,
said specialized image processing circuits providing electronic instructions to a non-transitory electronic storage hardware device to record images uploaded by a user over a computerized network to a social networking site;
said specialized image processing circuits providing electronic instructions to said non-transitory electronic storage hardware device to record categories of social networking boards to which said images are uploaded by said user to identify user image board posting behavior;
said specialized image processing circuits analyzing features within said images using automated image feature identification to identify user image posted content of each of said images;
said specialized image processing circuits determining user characteristics of said user based only on said user image board posting behavior and said user image posted content; and
said specialized image processing circuits outputting said user characteristics of said user on a graphic user interface hardware device.

14. The image processor device according to claim 13 , said analyzing said features within said images generating vectors describing said user image posted content.

15. The image processor device according to claim 13 , said analyzing said features within said images determining an occurrence count of each type of feature in each said image to produce a vector of each said image.

16. The image processor device according to claim 15 , said analyzing said features within said images determining vectors representing said features,
said image processor device further comprising employing a latent semantic index to discover topics making up a length of each said vector.

17. The image processor device according to claim 13 , said user characteristics comprising at least one of age, gender, preferences, political orientation, and interests.

18. The image processor device according to claim 13 , further comprising identifying said features in said images by employing scale-invariant feature transform (SIFT) to said images.

Assignees

• Xerox Corp

Inventors

• You Quanzeng
• Bhatia Sumit

Classifications

• G06V20/70

  Labelling scene content, e.g. deriving syntactic or semantic representations · CPC title

• G06V20/30 Primary

  in albums, collections or shared content, e.g. social network photos or video · CPC title

• G06V10/464

  using a plurality of salient features, e.g. bag-of-words [BoW] representations · CPC title

• G06K9/4676 Primary

  Physics · mapped topic

• G06K9/00677

  Physics · mapped topic