Recommendation agent using a personality model determined from mobile device data

• US 9,020,864 B2
• Filed: 07/24/2013
• Issued: 04/28/2015
• Est. Priority Date: 07/25/2012
• Status: Active Grant

• Current Assignee: ARO, Inc. (Dibona Holding)
• Original Assignee: ARO, Inc. (Dibona Holding)
• Inventors: Perkowitz, Michael, Eustice, Kevin Francis, Hickl, Andrew F.
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): Nilsson, Eric
• Application Number:
  US13/950,117
• Publication Number:
  US 20140032452A1
• Time in Patent Office: 643 Days
• Field of Search: None
• US Class Current: 706/12
• CPC Class Codes: G06F 16/955 using information identifiers, e.g. uniform resource locators [URL]
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: H04W 4/02 Services making use of location information
• CPC Class Codes: H04W 4/029 Location-based management or tracking services
• CPC Class Codes: H04W 4/21 for social networking applications

Abstract

A user'"'"'s context history is analyzed to build a personality model describing the user'"'"'s personality and interests. The personality model includes a plurality of metrics indicating the user'"'"'s position on a plurality of personality dimensions, such as desire for novelty, tendency for extravagance, willingness to travel, love of the outdoors, preference for physical activity, and desire for solitude. A customized recommendation agent is then built based on the personality model, which selects a recommendation from a corpus to present to the user based on an affinity between the user'"'"'s personality and the selected recommendation.

Claims

1. A method of creating a customized recommendation agent for a user, the method comprising:

obtaining a plurality of labelled context slices derived from context data associated with a user, each labelled context slice including a time, a location and a user context label specifying at least a place inferred from the location;

obtaining place features of places included in the obtained plurality of labelled context slices, the obtained place features relevant to personality traits of the user;

identifying, using the plurality of labelled context slices, one or more home areas corresponding to one or more places at which the user has spent a majority of time spanned by the labelled context slices;

identifying, from the places included in the plurality of labelled context slices, non-home areas comprising places that do not correspond to a home area;

determining a home area statistic and non-home area statistics from the obtained place features, the home area statistic describing place features of the one or more home areas, the non-home area statistics describing place features of the non-home areas;

determining, by a processor, a plurality of personality metrics based on the home area statistic and the non-home area statistics, each personality metric quantifying a position of the user on a corresponding one of a plurality of personality trait dimensions, wherein determining the plurality of personality metrics comprises applying a machine learning algorithm to the places to determine the plurality of personality metrics, the machine learning algorithm trained by;

obtaining personality scores of each user of a baseline group for the plurality of metrics;

obtaining baseline contextual slices for each user of the baseline group, the baseline contextual slices derived from context data associated with the user of the baseline group, the baseline contextual slices including locations and baseline contextual labels specifying places inferred from the locations; and

training the machine learning algorithm to predict the personality scores using the baseline contextual labels from the baseline contextual slices obtained for each user; and

creating the customized recommendation agent configured to provide a recommendation to the user responsive to the plurality of personality metrics indicating the user is likely to find value in the recommendation.

2. The method of claim 1, wherein the plurality of personality trait dimensions include at least two of:

desire for novelty, desire for extravagance, willingness to travel, love of the outdoors, preference for physical activity, and desire for solitude.

3. The method of claim 1, wherein providing the recommendation further comprises:

identifying a reason why the recommendation was selected; and

providing the reason for presentation to the user in conjunction with the recommendation.

4. The method of claim 1, wherein providing the recommendation to the user comprises:

receiving an input context associated with the user;

selecting the recommendation from a corpus of recommendations based on the plurality of personality metrics and the input context; and

providing the recommendation for presentation to the user.

5. The method of claim 4, wherein selecting the recommendation comprises:

calculating a weight for each of a plurality of recommendations from the corpus, each weight based on a degree of correspondence between the plurality of personality metrics and a corresponding recommendation; and

selecting the recommendation from the plurality of recommendations responsive to the weight corresponding to the recommendation.

6. The method of claim 5, wherein the recommendation corresponds to a venue, the method further comprising:

adjusting the weight corresponding to the recommendation based on a rating for the venue provided by another user.

7. The method of claim 1, further comprising:

receiving feedback indicating how the user responded to the recommendation; and

updating at least one personality metric of the plurality based on the feedback.

8. The method of claim 7, wherein the feedback indicates one of:

the user following the recommendation, the user adding the recommendations to a plan, the user partially following the recommendation, or the user rejecting the recommendation.

9. The method of claim 1, further comprising:

providing the user with a series of questions, each question a binary choice that determines affinity for one of the personality trait dimensions; and

adjusting at least one personality metric of the plurality based on the user'"'"'s responses to the series of questions.

10. The method of claim 1,

wherein obtaining the place features comprises determining categories that group similar places, and

wherein determining the non-home area statistics comprises determining frequencies of visits to different categories of the non-home areas.

11. The method of claim 1,

wherein obtaining the place features comprises determining distances from each of the non-home areas to a geographically nearest home area, and

wherein determining the non-home area statistics comprises determining a non-home area statistic summarizing the determined distances from each of the non-home areas to a geographically nearest home area.

12. The method of claim 1, wherein determining the home area statistic comprises determining a proportion of visits to one of the home areas relative to total visits to the places from the plurality of labelled context slices.

13. A non-transitory computer-readable storage medium comprising executable computer program code, the computer program code comprising instructions for:

obtaining a plurality of labelled context slices derived from context data associated with a user, each labelled context slice including a time, a location and a user context label specifying at least a place inferred from the location;

obtaining place features of places included in the obtained plurality of labelled context slices, the obtained place features relevant to personality traits of the user;

identifying, using the plurality of labelled context slices, one or more home areas corresponding to one or more places at which the user has spent a majority of time spanned by the labelled context slices;

identifying, from the places included in the plurality of labelled context slices, non-home areas comprising places that do not correspond to a home area;

determining a home area statistic and non-home area statistics from the obtained place features, the home area statistic describing place features of the one or more home areas, the non-home area statistics describing place features of the non-home areas;

determining, by a processor, a plurality of personality metrics based on the home area statistic and the non-home area statistics, each personality metric quantifying a position of the user on a corresponding one of a plurality of personality trait dimensions, wherein determining the plurality of personality metrics comprises applying a machine learning algorithm to the places to determine the plurality of personality metrics, the machine learning algorithm trained by;

obtaining personality scores of each user of a baseline group for the plurality of metrics;

obtaining baseline contextual slices for each user of the baseline group, the baseline contextual slices derived from context data associated with the user of the baseline group, the baseline contextual slices including locations and baseline contextual labels specifying places inferred from the locations; and

training the machine learning algorithm to predict the personality scores using the baseline contextual labels from the baseline contextual slices obtained for each user; and

creating a customized recommendation agent configured to provide a recommendation to the user responsive to the plurality of personality metrics indicating the user is likely to find value in the recommendation.

14. The non-transitory computer-readable storage medium of claim 13, wherein the computer program code further comprises instructions for:

identifying a reason why the recommendation was selected; and

providing the reason to the user in conjunction with the recommendation.

15. The non-transitory computer-readable storage medium of claim 13, wherein selecting the recommendation comprises:

receiving an input context associated with the user;

selecting the recommendation from a corpus of recommendations based on the plurality of personality metrics and the input context;

calculating a weight for each of a plurality of recommendations from the corpus, each weight based on a degree of correspondence between the plurality of personality metrics and a corresponding recommendation; and

selecting the recommendation from the plurality of recommendations responsive to the weight corresponding to the recommendation; and

providing the recommendation for presentation to the user.

16. The non-transitory computer-readable storage medium of claim 15, wherein the recommendation corresponds to a venue, the method further comprising:

adjusting the weight corresponding to the recommendation based on a rating for the venue provided by another user.

17. The non-transitory computer-readable storage medium of claim 13, wherein the computer program code further comprises instructions for:

obtaining feedback indicating how the user responded to the recommendation; and

updating at least one personality metric of the plurality based on the feedback.

18. The non-transitory computer-readable storage medium of claim 13, wherein the computer program code further comprises instructions for:

providing the user with a series of questions, each question a binary choice that determines affinity for one of the personality trait dimensions; and

adjusting at least one personality metric of the plurality based on the user'"'"'s responses to the series of questions.