Granted Patent B2

US 11,620,472 · App. 16/856,430 · Granted Apr 4, 2023

Inventors: Samuel Christopher John Plant (Cambridge, GB); Nathan Alexander Burn (Cambridge, GB); John Matthew Dilley (Longstanton, GB); Ellen Rose Wootten (Bury St. Edmunds, GB); Nilpa Madhusudan Shah (Cambridge, GB)

Assignee: Citrix Systems, Inc.

G06K9/6256G06F9/542G06F16/24578G06F16/9535G06K9/6262G06N3/0445G06N20/00

Abstract

Claims (105)

1. A computer system for identifying individuals with a user-requested expertise related to one or more software applications, the system comprising:

a memory;

a network interface; and

at least one processor coupled to the memory and the network interface and configured to

receive, via the network interface, a user input from a user,

extract one or more keywords from the user input,

generate one or more search requests based upon the one or more keywords, each search request of the one or more search requests identifying at least one application programming interface (API) call configured to invoke at least one API function as exposed by a software application of the one or more software applications,

transmit the one or more search requests to the one or more software applications,

receive one or more search responses from the one or more software applications,

determine, from the one or more search responses, a plurality of software application users and a set of evidence associated with each software application user of the plurality of software application users, each set of evidence including a listing of user interactions with each of the one or more software applications,

aggregate the sets of evidence into an aggregated data set,

configure the aggregated data set as an input to a machine learning classifier,

input a set of initial evidence weightings into the machine learning classifier to generate a trained classifier, the set of initial evidence weightings comprising at least one weight parameter for the set of evidence,

input the aggregated data set into the trained classifier, and

receive a ranked listing of the plurality of software application users from the trained classifier, the ranked listing generated by the trained classifier based upon the aggregated data set.

2. The computer system of claim 1 , wherein the at least one processor is further configured to:

process the ranked listing to produce a personalized response to the user input; and

transmit, via the network interface, the personalized response to the user.

3. The computer system of claim 2 , wherein the at least one processor is further configured to:

receive user feedback from the user regarding the personalized response, the user feedback comprising a user feedback score of one or more of the plurality of software application users included in the personalized response;

generate an updated set of evidence weightings based upon the user feedback; and

store the updated set of evidence weightings.

4. The computer system of claim 3 , wherein the at least one processor is further configured to:

receive an updated user input from the user;

generate an updated aggregated data set based upon the updated user input;

input the updated set of evidence weightings into the trained classifier to generate a retrained classifier;

input the updated aggregated data set into the retrained classifier; and

receive an updated ranked listing of the plurality of software application users from the retrained classifier, the updated ranked listing generated by the retrained classifier based upon the updated aggregated data set.

5. The computer system of claim 2 , wherein the at least one processor is further configured to:

monitor user telemetry data indicative of one or more interactions between the user and the personalized response;

generate an updated set of evidence weightings based upon the user telemetry data; and

store the updated set of evidence weightings.

6. The computer system of claim 5 , wherein the at least one processor is further configured to:

receive an updated user input from the user;

generate an updated aggregated data set based upon the updated user input;

input the updated aggregated data set and the updated set of evidence weightings into the trained classifier; and

receive an updated ranked listing of the plurality of software application users from the trained classifier, the updated ranked listing generated by the trained classifier based upon the updated aggregated data set and the updated set of evidence weightings.

7. The computer system of claim 1 , wherein each of the listing of user interactions includes information related to the one or more keywords.

8. The computer system of claim 1 , wherein the one or more software applications comprise at least one of an email application, a chat application, a messaging application, a community forum application, a code generation and/or review application, an accounting application, and a word processing application.

9. A method of identifying individuals with a user-requested expertise related to one or more software applications, the method comprising:

receiving, by at least one processor, a user input from a user;

extracting, by the at least one processor, one or more keywords from the user input;

generating, by the at least one processor, one or more search requests based upon the one or more keywords, each search request of the one or more search requests identifying at least one application programming interface (API) call configured to invoke at least one API function as exposed by a software application of the one or more software applications;

transmitting, by the at least one processor, the one or more search requests to the one or more software applications;

receiving, by the at least one processor, one or more search responses from the one or more software applications;

determining, by the at least one processor, a plurality of software application users and a set of evidence associated with each software application user of the plurality of software application users from the one or more search responses, each set of evidence including a listing of user interactions with each of the one or more software applications;

aggregating, by the at least one processor, the sets of evidence into an aggregated data set;

configuring, by the at least one processor, the aggregated data set as an input to a machine learning classifier;

inputting, by the at least one processor, a set of initial evidence weightings into the machine learning classifier to generate a trained classifier, the set of initial evidence weightings comprising at least one weight parameter for the set of evidence,

inputting, by the at least one processor, the aggregated data set into the trained classifier; and

receiving, by the at least one processor, a ranked listing of the plurality of software application users from the trained classifier, the ranked listing generated by the trained classifier based upon the aggregated data set.

10. The method of claim 9 , further comprising:

processing, by the at least one processor, the ranked listing to produce a personalized response to the user input; and

transmitting, by the at least one processor, the personalized response to the user.

11. The method of claim 10 , further comprising:

receiving, by the at least one processor, user feedback from the user regarding the personalized response, the user feedback comprising a user feedback score of one or more of the plurality of software application users included in the personalized response;

generating, by the at least one processor, an updated set of evidence weightings based upon the user feedback;

inputting, by the at least one processor, the updated set of evidence weightings into the trained classifier to generate a retrained classifier;

receiving, by the at least one processor, an updated user input from the user;

generating, by the at least one processor, an updated aggregated data set based upon the updated user input;

inputting, by the at least one processor, the updated aggregated data set into the retrained classifier; and

receiving, by the at least one processor, an updated ranked listing of the plurality of software application users from the retrained classifier, the updated ranked listing generated by the retrained classifier based upon the updated aggregated data set.

12. The method of claim 10 , further comprising:

monitoring, by the at least one processor, user telemetry data indicative of one or more interactions between the user and the personalized response;

generating, by the at least one processor, an updated set of evidence weightings based upon the user telemetry data;

storing, by the at least one processor, the updated set of evidence weightings;

receiving, by the at least one processor, an updated user input from the user;

generating, by the at least one processor, an updated aggregated data set based upon the updated user input;

inputting, by the at least one processor, the updated aggregated data set and the updated set of evidence weightings into the trained classifier; and

receiving, by the at least one processor, an updated ranked listing of the plurality of software application users from the trained classifier, the updated ranked listing generated by the trained classifier based upon the updated aggregated data set and the updated set of evidence weightings.

13. The method of claim 9 , wherein each of the listing of user interactions includes information related to the one or more keywords.

14. A non-transitory computer readable medium storing computer executable instructions to identify individuals with a user-requested expertise related to one or more software applications, the computer executable instructions comprising instructions to:

receive a user input from a user;

extract one or more keywords from the user input;

generate one or more search requests based upon the one or more keywords, each search request of the one or more search requests identifying at least one application programming interface (API) call configured to invoke at least one API function as exposed by a software application of the one or more software applications;

transmit the one or more search requests to the one or more software applications;

receive one or more search responses from the one or more software applications;

determine, from the one or more search responses, a plurality of software application users and a set of evidence associated with each software application user of the plurality of software application users, each set of evidence including a listing of user interactions with each of the one or more software applications;

aggregate the sets of evidence into an aggregated data set; configure the aggregated data set as an input to a machine learning classifier;

input a set of initial evidence weightings to the machine learning classifier to generate a trained classifier, the set of initial evidence weightings comprising at least one weight parameter for the set of evidence,

input the aggregated data set into the trained classifier; and

receive a ranked listing of the plurality of software application users from the trained classifier, the ranked listing generated by the trained classifier based upon the aggregated data set.

15. The computer readable medium of claim 14 , wherein the instructions further comprise instructions to:

process the ranked listing to produce a personalized response to the user input; and

transmit the personalized response to the user.

16. The computer readable medium of claim 15 , wherein the instructions further comprise instructions to:

receive user feedback from the user regarding the personalized response, the user feedback comprising a user feedback score of one or more of the plurality of software application users included in the personalized response;

generate an updated set of evidence weightings based upon the user feedback;

input the updated set of evidence weightings into the trained classifier to generate a retrained classifier;

receive an updated user input from the user;

generate an updated aggregated data set based upon the updated user input;

input the updated aggregated data set into the retrained classifier; and

receive an updated ranked listing of the plurality of software application users from the retrained classifier, the updated ranked listing generated by the retrained classifier based upon the updated aggregated data set.

17. The computer readable medium of claim 15 , wherein the instructions further comprise instructions to:

monitor user telemetry data indicative of one or more interactions between the user and the personalized response;

generate an updated set of evidence weightings based upon the user telemetry data;

store the updated set of evidence weightings;

receive an updated user input from the user;

generate an updated aggregated data set based upon the updated user input;

input the updated aggregated data set and the updated set of evidence weightings into the trained classifier; and

receive an updated ranked listing of the plurality of software application users from the trained classifier, the updated ranked listing generated by the trained classifier based upon the updated aggregated data set and the updated set of evidence weightings.

18. The computer readable medium of claim 14 , wherein each of the listing of user interactions includes information related to the one or more keywords.

19. The computer readable medium of claim 14 , wherein the set of initial evidence weightings is selected based upon the one or more keywords.

20. The computer system of claim 1 , wherein the set of initial evidence weightings is selected based upon the one or more keywords.

21. The method of claim 9 , further comprising selecting the set of initial evidence weightings based upon the one or more keywords.

Assignments (9)