Performing Cognitive Operations Based on an Aggregate User Model of Personality Traits of Users

• US 20160170967A1
• Filed: 12/11/2014
• Published: 06/16/2016
• Est. Priority Date: 12/11/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Rodriguez, Laura J., Allen, Corville O.
• Granted Patent: US 10,090,002 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 40/279 Recognition of textual entities
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 25/63 for estimating an emotional state

Abstract

Mechanisms, in a natural language processing (NLP) system are provided. The NLP system receives a plurality of communications associated with a communication system, over a predetermined time period, from a plurality of end user devices. The NLP system identifies, for each communication in the plurality of communications, a user submitting the communication to thereby generate a set of users comprising a plurality of users associated with the plurality of communications. The NLP system retrieves a user model for each user in the set of users, which specifies at least one attribute of a corresponding user. The NLP system generates an aggregate user model that aggregates the at least one attribute of each user in the set of users together to generate an aggregate representation of the attributes of the plurality of users in the set of users. The NLP system performs a cognitive operation based on the aggregate user model.

Claims

1. A method, in a natural language processing (NLP) system comprising a processor and a memory, the method comprising:

receiving, by the NLP system, a plurality of communications associated with a communication system, over a predetermined time period, from a plurality of end user devices;

identifying, by the NLP system, for each communication in the plurality of communications, a user submitting the communication to thereby generate a set of users comprising a plurality of users associated with the plurality of communications;

retrieving, by the NLP system, a user model for each user in the set of users, wherein the user model specifies at least one attribute of a corresponding user;

generating, by the NLP system, an aggregate user model that aggregates the at least one attribute of each user in the set of users together to generate an aggregate representation of the attributes of the plurality of users in the set of users; and

performing, by the NLP system, a cognitive operation based on the aggregate user model.

2. The method of claim 1, wherein, for each user in the set of users, the at least one attribute comprises at least one personality trait of the user.

3. The method of claim 2, wherein retrieving the user model for each user in the set of users comprises:

determining the at least one personality trait associated with the user, and wherein the at least one personality trait of the user is identified based on at least one of information specified in a stored profile of the user or indicators in natural language of one or more communications, in the plurality of communications, associated with the user as identified by natural language processing performed on the one or more communications.

4. The method of claim 3, wherein the stored profile of the user is a user profile stored in association with a social networking website system through which the plurality of communications are received by the NLP system.

5. The method of claim 2, wherein generating the aggregate user model comprises at least one of combining a like personality trait, of the set of users, that are within a standard deviation of a variance of the like personality trait, or combining personality traits, of the set of users, that are within a fixed range for a majority of the users in the set of users.

6. The method of claim 2, wherein generating the aggregate user model comprises weighting each personality trait of each user in the set of users using a mean square weighted deviation.

7. The method of claim 2, wherein generating the aggregate user model comprises:

correlating personality trait values, for each personality trait specified in the user models of the set of users, to identify correlated personality trait values from among the user models;

eliminating non-correlating personality trait values that are more than one standard deviation away from other similar personality trait values; and

summing, for each personality trait specified in the user models, the correlated personality trait values to generate an aggregate score for each personality trait.

8. The method of claim 7, wherein correlating personality trait values comprises:

calculating a standard deviation for all personality trait values specified in the user models for a selected personality trait;

generating a plurality of brackets of personality trait values for the personality trait, wherein each bracket in the plurality of brackets represents a separate and distinct range of personality trait values from the other brackets in the plurality of brackets; and

associating each personality trait value in the personality trait values specified in the user models with one of the brackets in the plurality of brackets.

9. The method of claim 8, wherein generating the aggregate user model further comprises:

for each bracket in the plurality of brackets, counting a number of personality trait values falling within the bracket to generate a bracket count; and

selecting at least one bracket of the plurality of brackets having a largest bracket count.

10. The method of claim 8, wherein generating the aggregate user model further comprises:

for each bracket in the plurality of brackets, counting a number of personality trait values falling within the bracket to generate a bracket count;

selecting two brackets from the plurality of brackets based on the bracket counts of the plurality of brackets;

determining if the selected brackets have ranges of personality trait values that are adjacent to each other; and

in response to the selected brackets having ranges of personality trait values that are adjacent to each other, selecting the personality trait values that are within the selected brackets and eliminating the lowest and highest personality trait values that are one standard deviation away from high and low personality trait values of the selected brackets.

11. The method of claim 1, wherein the plurality of communications are communications directed to a same topic exchanged by the set of users within the predetermined time period.

12. The method of claim 1, wherein the cognitive operation comprises at least one of forwarding a suggestion specifies in one or more of the communications in the plurality of communications to another user based on a matching of a user model of the other user being associated with the aggregate user model, scoring a candidate answer to a question based on a correlation of the user model of the other user being associated with the aggregate user model, or filtering messages exchanged during the predetermined time period.

13. A computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed on a computing device implementing a natural language processing (NLP) system, causes the computing device to:

receive, by the NLP system, a plurality of communications associated with a communication system, over a predetermined time period, from a plurality of end user devices;

identify, by the NLP system, for each communication in the plurality of communications, a user submitting the communication to thereby generate a set of users comprising a plurality of users associated with the plurality of communications;

retrieve, by the NLP system, a user model for each user in the set of users, wherein the user model specifies at least one attribute of a corresponding user;

generate, by the NLP system, an aggregate user model that aggregates the at least one attribute of each user in the set of users together to generate an aggregate representation of the attributes of the plurality of users in the set of users; and

perform, by the NLP system, a cognitive operation based on the aggregate user model.

14. The computer program product of claim 13, wherein, for each user in the set of users, the at least one attribute comprises at least one personality trait of the user.

15. The computer program product of claim 14, wherein the computer readable program further causes the computing device to retrieve the user model for each user in the set of users at least by:

determining the at least one personality trait associated with the user, and wherein the at least one personality trait of the user is identified based on at least one of information specified in a stored profile of the user or indicators in natural language of one or more communications, in the plurality of communications, associated with the user as identified by natural language processing performed on the one or more communications.

16. The computer program product of claim 15, wherein the stored profile of the user is a user profile stored in association with a social networking website system through which the plurality of communications are received by the NLP system.

17. The computer program product of claim 14, wherein the computer readable program further causes the computing device to generate the aggregate user model at least by at least one of combining a like personality trait, of the set of users, that are within a standard deviation of a variance of the like personality trait, or combining personality traits, of the set of users, that are within a fixed range for a majority of the users in the set of users.

18. The computer program product of claim 14, wherein the computer readable program further causes the computing device to generate the aggregate user model at least by weighting each personality trait of each user in the set of users using a mean square weighted deviation.

19. The computer program product of claim 14, wherein the computer readable program further causes the computing device to generate the aggregate user model at least by:

correlating personality trait values, for each personality trait specified in the user models of the set of users, to identify correlated personality trait values from among the user models;

eliminating non-correlating personality trait values that are more than one standard deviation away from other similar personality trait values; and

summing, for each personality trait specified in the user models, the correlated personality trait values to generate an aggregate score for each personality trait.

20. The computer program product of claim 19, wherein correlating personality trait values comprises:

calculating a standard deviation for all personality trait values specified in the user models for a selected personality trait;

generating a plurality of brackets of personality trait values for the personality trait, wherein each bracket in the plurality of brackets represents a separate and distinct range of personality trait values from the other brackets in the plurality of brackets; and

associating each personality trait value in the personality trait values specified in the user models with one of the brackets in the plurality of brackets.

21. The computer program product of claim 20, wherein the computer readable program further causes the computing device to generate the aggregate user model at least by:

for each bracket in the plurality of brackets, counting a number of personality trait values falling within the bracket to generate a bracket count; and

selecting at least one bracket of the plurality of brackets having a largest bracket count.

22. The computer program product of claim 20, wherein the computer readable program further causes the computing device to generate the aggregate user model at least by:

for each bracket in the plurality of brackets, counting a number of personality trait values falling within the bracket to generate a bracket count;

selecting two brackets from the plurality of brackets based on the bracket counts of the plurality of brackets;

determining if the selected brackets have ranges of personality trait values that are adjacent to each other; and

in response to the selected brackets having ranges of personality trait values that are adjacent to each other, selecting the personality trait values that are within the selected brackets and eliminating the lowest and highest personality trait values that are one standard deviation away from high and low personality trait values of the selected brackets.

23. The computer program product of claim 13, wherein the plurality of communications are communications directed to a same topic exchanged by the set of users within the predetermined time period.

24. The computer program product of claim 13, wherein the cognitive operation comprises at least one of forwarding a suggestion specifies in one or more of the communications in the plurality of communications to another user based on a matching of a user model of the other user being associated with the aggregate user model, scoring a candidate answer to a question based on a correlation of the user model of the other user being associated with the aggregate user model, or filtering messages exchanged during the predetermined time period.

25. An apparatus comprising:

a processor; and

a memory coupled to the processor, wherein the memory comprises instructions which, when executed by the processor, cause the processor to;

receive a plurality of communications associated with a communication system, over a predetermined time period, from a plurality of end user devices;

identify for each communication in the plurality of communications, a user submitting the communication to thereby generate a set of users comprising a plurality of users associated with the plurality of communications;

retrieve a user model for each user in the set of users, wherein the user model specifies at least one attribute of a corresponding user;

generate an aggregate user model that aggregates the at least one attribute of each user in the set of users together to generate an aggregate representation of the attributes of the plurality of users in the set of users; and

perform a cognitive operation based on the aggregate user model.