Collecting naturally expressed affective responses for training an emotional response predictor utilizing voting on content

• US 9,224,175 B2
• Filed: 10/23/2014
• Issued: 12/29/2015
• Est. Priority Date: 10/14/2012
• Status: Active Grant

• Current Assignee: IP3 2017, Series 200 of Allied Security Trust I (Allied Security Trust)
• Original Assignee: Ari M. Frank, Gil Thieberger
• Inventors: Frank, Ari M, Thieberger, Gil
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): LAMARDO, VIKER ALEJANDRO
• Application Number:
  US14/521,894
• Publication Number:
  US 20150039405A1
• Time in Patent Office: 432 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: G07C 13/00 Voting apparatus
• CPC Class Codes: H04N 21/44218 Detecting physical presence or behaviour of the user, e.g. using sensors to detect if the user is leaving the room or changes his face expression during a TV program methods or arrangements for recognising human body or animal bodies or body parts G06V40/10; methods or arrangements for acquiring or recognising human faces, facial parts, facial sketches, facial expressions G06V40/16; methods or arrangements for recognising movements or behaviour G06V40/20; arrangements for identifying users in broadcast systems H04H60/45

Abstract

Described herein are systems, methods, and computer program products for collecting naturally expressed affective responses for training an emotional response predictor utilizing voting on content. In one embodiment, a label generator is configured to receive a vote, provided by a user, on a segment of content consumed by the user. The label generator determines whether the user consumed the segment during a duration that is shorter than a predetermined threshold, and utilizes the vote to generate a label related to an emotional response to the segment. A training module receives the label and measurement of an affective response of the user taken, at least in part, during a period that starts at most 30 seconds before the vote is provided, and trains the emotional response predictor with the measurement and the label.

Claims

1. A system configured to collect naturally expressed affective responses for training an emotional response predictor utilizing voting on content, comprising:

a voting mechanism configured to receive a vote from a user upon the user consuming a segment of content;

wherein the user provides the vote via one or more of the following mechanisms;

a like voting mechanism, a dislike voting mechanism, a star rating mechanism, a numerical rating mechanism, an up voting mechanism, a down voting mechanism, a ranking mechanism, and a mechanism for entering comments comprising data representing text;

a label generator configured to receive the vote and an indication of a duration during which the user consumed the segment of content;

the label generator is further configured to;

determine that the duration is shorter than a predetermined threshold, and utilize the vote to generate a label related to an emotional response to the segment; and

a training module configured to receive the label and an affective response measurement of the user taken with a sensor coupled to the user while the user consumed the segment of content, and to train the emotional response predictor with the measurement and the label;

wherein the measurement is taken, at least in part, during a period that starts at most 30 seconds before the vote is provided.

2. The system of claim 1, wherein the voting mechanism is a voting mechanism belonging to a social network, and the predetermined threshold is less than 15 seconds.

3. The system of claim 1, wherein the voting mechanism is offered independently of whether the vote is used to trigger a sensor to acquire the naturally expressed affective responses.

4. The system of claim 3, wherein the period at which the affective response measurement is taken ends at most 10 seconds after the vote is provided.

5. The system of claim 3, wherein the voting mechanism is configured to enable the user to enter a comment comprising data representing text;

and wherein the label generator is further configured to perform semantic analysis on the text in order to generate the label.

6. The system of claim 1, wherein the system is configured not to forward to the training module a second affective response measurement that does not have a corresponding label generated from a vote.

7. The system of claim 1, wherein the label generator is further configured to suggest to the training module a training weight for the affective response measurement of the user.

8. The system of claim 1, wherein the label generator is further configured to receive data derived from an analysis of the segment of content, and to utilize the data to generate the label.

9. The system of claim 1, wherein the label generator is further configured to receive an estimated label based on affective responses of other users to the segment of content, and to utilize the estimated label to generate the label.

10. The system of claim 1, wherein at least one affective response measurement that has a corresponding label generated from a vote, and at least one affective response measurement that does not have a corresponding label generated from a vote, are affective response measurements of the user to content provided by the same friend of the user on a social network.

11. The system of claim 1, wherein the training module is further configured to apply a training weight to a first affective response measurement that has a corresponding label generated from a vote that is at least double the training weight applied to a second affective response measurement that does not have a corresponding label generated from a vote.

12. A method for collecting naturally expressed affective responses for training an emotional response predictor utilizing voting on content, comprising:

receiving a vote from a user upon the user consuming a segment of content;

wherein the user provides the vote via a voting mechanism; and

wherein the voting mechanism comprises one or more of the following;

a like voting mechanism, a dislike voting mechanism, a star rating mechanism, a numerical rating mechanism, an up voting mechanism, a down voting mechanism, a ranking mechanism, and a mechanism for entering comments comprising data representing text;

receiving an indication of a duration during which the user consumed the segment of content;

determining that the duration is shorter than a predetermined threshold, and utilizing the vote for generating a label related to an emotional response to the segment;

receiving an affective response measurement of the user taken with a sensor coupled to the user while the user consumed the segment of content;

wherein the measurement is taken, at least in part, during a period that starts at most 30 seconds before the vote is provided; and

training the emotional response predictor with the label and the affective response measurement.

13. The method of claim 12, wherein the voting mechanism is a voting mechanism belonging to a social network, and the voting mechanism is offered independently of whether the vote is used to trigger a sensor to acquire the affective response measurements.

14. The method of claim 13, wherein the period at which the affective response measurement is taken ends at most 10 second after the vote is made.

15. The method of claim 13, wherein the voting mechanism enables the user to enter a comment comprising data representing text, and generating the label involves performing semantic analysis on the text.

16. The method of claim 12, further comprising suggesting a training weight for the affective response measurement of the user.

17. The method of claim 12, wherein the training of the emotional response predictor further comprises applying a training weight to a first measurement that has a corresponding label generated from a vote that is at least double the training weight applied to a second measurement that does not have a corresponding label generated from a vote.

18. The method of claim 12, further comprising receiving data derived from an analysis of the content and utilizing the data to generate the label.

19. The method of claim 12, further comprising receiving a comment provided by the user comprising data representing text, and generating the label by performing semantic analysis on the text.

20. A non-transitory computer-readable medium for use in a computer to collect naturally expressed affective responses for training an emotional response predictor utilizing voting on content;

the computer comprises a processor, and the non-transitory computer-readable medium comprising;

program code for receiving a vote from a user upon the user consuming a segment of content;

wherein the user provides the vote via a voting mechanism; and

wherein the voting mechanism comprises one or more of the following;

a like voting mechanism, a dislike voting mechanism, a star rating mechanism, a numerical rating mechanism, an up voting mechanism, a down voting mechanism, a ranking mechanism, and a mechanism for entering comments comprising data representing text;

program code for receiving an indication of a duration during which the user consumed the segment of content;

program code for determining that the duration is shorter than a predetermined threshold, and for utilizing the vote for generating a label related to an emotional response to the segment;

program code for receiving an affective response measurement of the user taken with a sensor coupled to the user while the user consumed the segment of content;

wherein the measurement is taken, at least in part, during a period that starts at most 30 seconds before the vote is provided; and

program code for training the emotional response predictor with the label and the affective response measurement.