Software agents facilitating affective computing applications
Software agents collect measurements of affective response and provide them in a selective manner that maintains user privacy. In one embodiment, a sensor takes measurements of affective response of the user. A computer receives a request for an affective value indicative of an emotional response to an experience. The computer selects, from among the measurements, a certain measurement of affective response of the user that corresponds to an event in which the user had the experience. The computer then utilizes a model to calculate, based on the certain measurement, the affective value indicative of the emotional response of the user to experience. Optionally, the model is generated based on previously taken measurements of affective response of the user, taken while the user had various experiences, and indications of emotional responses the user had while said previously taken measurements were taken. The computer sends the affective value to fulfil the request.
1. A system for operating a software agent to assess and disclose crowd-based scores, comprising:
sensors configured to take measurements of affective response of users; and
a computer configured to:
calculate scores for experiences based on the measurements; wherein each score for an experience is calculated based on a subset of the measurements comprising measurements of at least five of the users who had the experience;
calculate a risk to privacy of a certain user due to a disclosure of one or more of the scores; wherein the certain user contributed a measurement of affective response to the calculation of each of the one or more of the scores; and wherein the risk is indicative of an extent of an improvement to an adversary bias model, which comprises a model of an Emotional Response Predictor (ERP), due to an update of the adversary bias model based on values of the one or more of the scores; and
disclose the one or more of the scores in a manner that is based on the risk; wherein responsive to the risk not reaching a threshold, the one or more of the scores are disclosed in a first manner, and responsive to the risk reaching the threshold, the one or more of the scores are disclosed in a second manner, which is less descriptive than the first manner.
2. The system of claim 1 , wherein the improvement is indicative of at least one of the following: (i) an increase, due to the update, in the accuracy of the ERP when utilizing the adversary bias model, and (ii) a decrease, due to the update, in a difference between the accuracy of the ERP when utilizing a model trained on measurements of affective response of the certain user and the accuracy of the ERP when utilizing the updated adversary bias model.
3. The system of claim 1 , wherein the computer is further configured to train the adversary bias model based on data comprising scores of experiences computed based on measurements corresponding to events, and descriptions of the events.
4. The system of claim 1 , wherein an experience involves doing at least one of the following: spending time at a certain location, consuming certain digital content, having a social interaction with a certain entity in the physical world, having a social interaction with a certain entity in a virtual world, viewing a certain live performance, performing a certain exercise, traveling a certain route, and consuming a certain product.
5. The system of claim 1 , wherein the computer is further configured to utilize an indication of a number of users who contributed measurements used to compute a score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby a larger number of users who contributed measurements reduces the risk to individual privacy.
6. The system of claim 1 , wherein the computer is further configured to utilize an indication of the variance of measurements used to compute a score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby a larger variance reduces the risk to individual privacy.
7. The system of claim 1 , wherein the computer is further configured to utilize an indication of a probability of a value of a disclosed score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby less probable scores increase the risk to individual privacy.
8. The system of claim 1 , wherein the computer is further configured to utilize an indication of a statistical significance of a disclosed score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby more significant scores increase the risk to individual privacy.
9. The system of claim 1 , wherein the computer is further configured to provide, when disclosing a score in the second manner, fewer values associated with the score, compared to the number of values associated with the score the computer is configured to provide when disclosing in the first manner.
10. The system of claim 1 , wherein the computer is further configured to provide, when disclosing a score in the second manner, fewer bits of information describing the score, compared to the number of bits of information describing the score the computer is configured to provide when disclosing in the first manner.
11. A method for operating a software agent to assess and disclose crowd-based scores, comprising:
receiving measurements of affective response of users;
calculating scores for experiences based on the measurements; wherein each score for an experience is calculated based on a subset of the measurements comprising measurements of at least five of the users who had the experience;
calculating a risk to privacy of a certain user due to a disclosure of one or more of the scores; wherein the certain user contributed a measurement of affective response to the calculation of each of the one or more of the scores; wherein the risk is indicative of an extent of an improvement to an adversary bias model, which comprises a model of an Emotional Response Predictor (ERP), due to an update of the adversary bias model based on values of the one or more of the scores; and
disclosing the one or more of the scores in a manner that is based on the risk; wherein responsive to the risk not reaching a threshold, the one or more of the scores are disclosed in a first manner, and responsive to the risk reaching the threshold, the one or more of the scores are disclosed in a second manner, which is less descriptive than the first manner.
12. The method of claim 11 , further comprising: calculating accuracy of the ERP when utilizing the updated adversary bias model, and detecting an increase in the accuracy corresponding to the improvement.
13. The method of claim 12 , further comprising: calculating a first accuracy of the ERP when utilizing a model trained on measurements of affective response of the certain user and a second accuracy of the ERP when utilizing the updated adversary bias model, and detecting a difference, between the first accuracy and the second accuracy, which corresponds to the improvement.
14. The method of claim 12 , further comprising: utilizing an indication of a number of users who contributed measurements used to compute a score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby a larger number of users who contributed measurements reduces the risk to individual privacy.
15. The method of claim 11 , further comprising: utilizing an indication of the variance of measurements used to compute a score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby a larger variance reduces the risk to individual privacy.
16. The method of claim 11 , further comprising: utilizing an indication of a probability of a value of a disclosed score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby less probable scores increase the risk to individual privacy.
17. The method of claim 11 , further comprising: utilizing an indication of a statistical significance of a disclosed score from among the one or more of the scores to calculate the risk to privacy of the certain user due to a disclosure of one or more of the scores; whereby more significant scores increase the risk to individual privacy.
18. The method of claim 11 , further comprising: providing, when disclosing a score in the second manner, fewer values associated with the score, compared to the number of values associated with the score provided when disclosing in the first manner.
19. The method of claim 11 , further comprising: providing, when disclosing a score in the second manner, fewer bits of information describing the score, compared to the number of bits of information describing the score provided when disclosing in the first manner.
20. A non-transitory computer-readable medium having instructions stored thereon that, in response to execution by a system including a processor and memory, causes the system to perform operations comprising:
receiving measurements of affective response of users;
calculating scores for experiences based on the measurements; wherein each score for an experience is calculated based on a subset of the measurements comprising measurements of at least five of the users who had the experience;
calculating a risk to privacy of a certain user due to a disclosure of one or more of the scores; wherein the certain user contributed a measurement of affective response to the calculation of each of the one or more of the scores; wherein the risk is indicative of an extent of an improvement to an adversary bias model, which comprises a model of an Emotional Response Predictor (ERP), due to an update of the adversary bias model based on values of the one or more of the scores; and
disclosing the one or more of the scores in a manner that is based on the risk; wherein responsive to the risk not reaching a threshold, the one or more of the scores are disclosed in a first manner, and responsive to the risk reaching the threshold, the one or more of the scores are disclosed in a second manner, which is less descriptive than the first manner.