Utilizing semantic analysis to determine how to measure affective response

• US 8,898,344 B2
• Filed: 03/27/2014
• Issued: 11/25/2014
• Est. Priority Date: 10/14/2012
• Status: Active Grant

• Current Assignee: Affectomatics ltd.
• Original Assignee: Ari M. Frank, Gil Thieberger
• Inventors: Thieberger, Gil, Frank, Ari M
• Primary Examiner(s): TSENG, CHENG YUAN
• Application Number:
  US14/227,020
• Publication Number:
  US 20140195221A1
• Time in Patent Office: 243 Days
• Field of Search: None
• US Class Current: 710/5
• CPC Class Codes: G06F 1/28 Supervision thereof, e.g. detecting power-supply failure by out of limits supervision
• CPC Class Codes: G06F 1/3206 Monitoring of events, devices or parameters that trigger a change in power modality
• CPC Class Codes: G06F 1/3212 Monitoring battery levels, e.g. power saving mode being initiated when battery voltage goes below a certain level
• CPC Class Codes: G06F 1/3234 Power saving characterised by the action undertaken
• CPC Class Codes: G06F 1/3296 by lowering the supply or operating voltage
• CPC Class Codes: G06F 2203/011 Emotion or mood input determined on the basis of sensed human body parameters such as pulse, heart rate or beat, temperature of skin, facial expressions, iris, voice pitch, brain activity patterns
• CPC Class Codes: G06F 3/013 Eye tracking input arrangements G06F3/015 takes precedence
• CPC Class Codes: G06F 3/015 Input arrangements based on nervous system activity detection, e.g. brain waves [EEG] detection, electromyograms [EMG] detection, electrodermal response detection
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 9/46 Multiprogramming arrangements
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

A semantic analyzer receives a segment of content, analyzes it utilizing semantic analysis, and outputs an indication regarding whether a value related to a predicted emotional response to the segment reaches a predetermined threshold. Based on the indication, a controller selects a measuring rate, from amongst at least first and second measuring rates, at which a device is to take measurements of affective response of a user to the segment. The first rate may be selected when the value does not reach the predetermined threshold, while the second mode may be selected when the value does reach it. The device takes significantly fewer measurements while operating at the first measuring rate, compared to number of measurements it takes while operating at the second measuring rate.

Claims

1. A system configured to use semantic analysis to measure affective response at varying measuring rates, comprising:

a semantic analyzer configured to receive a first segment of content comprising data representing first text;

the semantic analyzer is further configured to utilize semantic analysis to generate a first feature that describes an aspect of a meaning of a portion of the first text;

the semantic analyzer is further configured to generate, based on the first feature, a first indication that a first value related to a predicted emotional response to the first segment does not reach a first predetermined threshold; and

a controller configured to select, based on the first indication, a first measuring rate for a device for measuring affective response of a user to the first segment;

the semantic analyzer is further configured to receive a second segment of content comprising data representing second text;

the semantic analyzer is further configured to utilize semantic analysis to generate a second feature that describes an aspect of a meaning of a portion of the second text;

the semantic analyzer is further configured generate based on the second feature a second indication that a second value related to a predicted emotional response to the second segment does reach a second predetermined threshold;

the controller is further configured to select, based on the second indication, a second measuring rate for the device for measuring affective response of the user to the second segment;

wherein while operating at the first measuring rate, the device takes at least 50% fewer measurements of affective response, per unit of measurement time, compared to measurements of affective response the device takes while operating at the second measuring rate.

2. The system of claim 1, wherein the first and second predetermined thresholds represent first and second confidence levels in predictions of emotional response, respectively;

and wherein the first and second values represent confidence in predictions of emotional response to the first and second segments, respectively;

whereby the first measuring rate is selected when the first indication indicates that confidence in a prediction of emotional response to the first segment does not reach the first confidence level, and the second measuring rate is selected when the second indication indicates that confidence in a prediction of emotional response to the second segment does reach the second confidence level.

3. The system of claim 1, wherein the first and second predetermined thresholds represent first and second magnitudes of change in affective response, respectively;

and wherein the first and second values represent predicted changes in affective response due to exposure to the first and second segments, respectively;

whereby the first measuring rate is selected when the first indication indicates that a predicted change in affective response due to exposure to the first segment does not reach the first magnitude, and the second measuring rate is selected when the second indication indicates that a predicted change in affective response due to exposure to the second segment does reach the second magnitude.

4. The system of claim 1, wherein the first and second predetermined thresholds represent first and second benefit levels, respectively;

and wherein the first and second values represent predicted benefits to measuring affective response to the first and second segments, respectively;

whereby the first measuring rate is selected when the first indication indicates that a predicted benefit of measuring affective response to the first segment does not reach the first benefit level, and the second measuring rate is selected when the second indication indicates that a benefit of measuring affective response to the second segment does reach the second benefit level.

5. The system of claim 1, wherein when operating at the first measuring rate the device does not take affective response measurements of the user.

6. The system of claim 1, wherein the device is one or more of the following:

a physiological sensor, and an image capturing device.

7. The system of claim 1, further comprising a module configured to train an emotional response predictor (measurement ERP) based on feature values that are a product of measurements of the affective response of the user to the second segment of content and a label based on the second indication generated by the semantic analyzer.

8. The system of claim 1, wherein the semantic analyzer applies statistics on words that are correlated with certain emotional responses.

9. The system of claim 1, wherein the semantic analyzer applies Latent Semantic Analysis in order to associate a segment of content with a subject.

10. The system of claim 1, wherein the semantic analyzer evaluates the second segment shortly before the user is exposed to the second segment, whereby shortly is less than one minute.

11. The system of claim 1, wherein the second segment is received from an interactive computer game that provides context information about the second segment that may be utilized by the semantic analyzer to compute the second indication.

12. A method for utilizing semantic analysis to set a measuring rate of a device that measures affective response, comprising:

receiving a first indication derived from a first evaluation comprising semantic analysis of a first segment of content comprising data representing text;

wherein the first evaluation comprises analysis of an aspect of a meaning of a portion of the first text;

determining that the first indication indicates that a first value related to a prediction of emotional response to the first segment does not reach a first predetermined threshold, and configuring the device to measure, at a first measuring rate, affective response of a user to the first segment;

receiving a second indication derived from a second evaluation comprising semantic analysis of a second segment of content comprising data representing text;

wherein the second evaluation comprises analysis of an aspect of a meaning of a portion of the second text; and

determining that the second indication indicates that a second value related to a prediction of emotional response to the second segment reaches a second predetermined threshold, and configuring the device to measure, at a second measuring rate, affective response of the user to the second segment;

wherein while operating at the first measuring rate, the device takes at least 50% fewer measurements of affective response, per unit of measurement time, compared to measurements of affective response the device takes while operating at the second measuring rate.

13. The method of claim 12, wherein the semantic analysis of the data derived from the second segment is performed shortly before the user is exposed to the second segment, whereby shortly is less than one minute.

14. The method of claim 12, wherein the first and second indications indicate confidence in predictions of emotional response to the first and second segments, respectively;

whereby based on the first indication, it is determined that confidence in the prediction of the emotional response to the first segment does not reach the first predetermined threshold, and based on the second indication, it is determined that confidence in the prediction of the emotional response to the second segment does reach the second predetermined threshold.

15. The method of claim 12, wherein the first and second indications indicate expected magnitudes of change in affective response due to exposure to the first and second segments, respectively;

whereby based on the first indication, it is determined that expected magnitude of change in affective response due to exposure to the first segment does not reach the first predetermined threshold, and based on the second indication, it is determined that expected magnitude of change to affective response due to exposure to the second segment does reach the second predetermined threshold.

16. The method of claim 12, wherein the first and second indications indicate how beneficial it is to measure the affective response to the first and second segments, respectively;

whereby based on the first indication, it is determined that benefit from measuring the affective response to the first segment does not reach the first predetermined threshold, and based on the second indication it is determined that benefit from measuring affective response to the second segment does reach the second predetermined threshold.

17. The method of claim 12, wherein at the first measuring rate no measuring takes place of the affective response of the user to being exposed to the first segment.

18. The method of claim 12, wherein the device comprises a battery powered physiological sensor, and measuring the affective response of the user to the second segment comprises utilizing the device to take a physiological measurement of the user.

19. The method of claim 12, wherein the device is part of a battery powered image capturing device, and measuring the affective response of the user to the second segment comprises utilizing the image capturing device to capture an image of the user.

20. A system configured to enable a device to measure affective response at varying measuring rates based on semantic analysis of content, comprising:

a receiver configured to receive first and second indications derived from first and second evaluations comprising semantic analysis of first and second segments of content that comprise first and second data representing first and second texts, respectively;

wherein the first and second evaluations comprise analysis of first and second aspects of meanings of first and second portions of the first and second texts, respectively;

and wherein the first indication indicates that a first value, which is a function of emotional response the first segment, does not reach a first predetermined threshold, and the second indication indicates that a second value, which is a function of emotional response to the second segment, does reach a second predetermined threshold;

a processor configured to select, based on the first and second indications, first and second measuring rates for measuring affective response, respectively; and

a controller operable to configure the device to measure the affective response of a user according to the first and second measuring rates;

wherein while operating at the first measuring rate, the device takes at least 50% fewer measurements to measure affective response of the user to the first segment, per unit of measurement time, compared to measurements the device takes while operating at the second measuring rate to measure affective response of the user to the second segment.