Training a cognitive agent using document output generated from a recorded process

• US 9,799,326 B2
• Filed: 01/26/2016
• Issued: 10/24/2017
• Est. Priority Date: 01/26/2016
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Dhoolia, Pankaj, Gantayat, Neelamadhav, Gupta, Monika, Mani, Senthil Kumar Kumarasamy, Sinha, Vibha Singhal
• Primary Examiner(s): Chawan, Vijay B
• Application Number:
  US15/006,771
• Publication Number:
  US 20170213544A1
• Time in Patent Office: 637 Days
• Field of Search: 715716, 715273, 434365, 434350, 434262, 704251, 704270, 704275, 7042701, 704231
• CPC Class Codes: G06F 40/169 Annotation, e.g. comment data or footnotes
• CPC Class Codes: G06F 40/258 Heading extraction; Automatic titling; Numbering
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/0638 Interactive procedures

Abstract

One embodiment provides a method for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the method comprising: utilizing at least one processor to execute computer code that performs the steps of: obtaining a recording of a process, wherein the recording comprises a demonstration of executing the process; generating, using the recording, the process learning graph, wherein the process learning graph identifies a process flow; generating, using the recording, the document output, wherein the document output comprises process screen transitions and process steps; and providing the process learning graph and the document output to the cognitive agent. Other aspects are described and claimed.

Claims

1. A method for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the method comprising:

utilizing at least one processor to execute computer code that performs the steps of;

obtaining a recording of a process, wherein the recording comprises a demonstration of executing the process;

generating, using the recording, the process learning graph, wherein the process learning graph identifies a process flow;

generating, using the recording, the document output, wherein the document output comprises process screen transitions and process steps;

expanding a knowledge base of the cognitive agent by providing the process learning graph and the document output to the cognitive agent;

accessing, by the cognitive agent and responsive to receiving a request by a user corresponding to the process, the knowledge base for the process learning graph and the document output; and

assisting, using the cognitive agent, the user with at least one step of the process by traversing the process learning graph and confirming process flow steps.

2. The method of claim 1, wherein the process learning graph defines user input requirements.

3. The method of claim 2, wherein the process learning graph identifies format constraints for the user input requirements.

4. The method of claim 1, wherein the recording comprises a script.

5. The method of claim 1, wherein the recording comprises a video recording.

6. The method of claim 5, wherein the video recording comprises annotations identifying at least one of:

user inputs, screen headings, and actions.

7. The method of claim 6, comprising extracting the process flow, the process screen transitions, and the process steps using the annotations.

8. The method of claim 1, wherein the document output comprises instructions for steps to be completed on a process screen.

9. The method of claim 1, comprising identifying system responses based upon identified user inputs and wherein the document output comprises the system responses.

10. An apparatus for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the apparatus comprising:

at least one processor; and

a computer readable storage medium having computer readable program code embodied therewith and executable by the at least one processor, the computer readable program code comprising;

computer readable program code that obtains a recording of a process, wherein the recording comprises a demonstration of executing the process;

computer readable program code that generates, using the recording, the process learning graph, wherein the process learning graph identifies a process flow;

computer readable program code that generates, using the recording, the document output, wherein the document output comprises process screen transitions and process steps;

computer readable program code that expands a knowledge base of the cognitive agent by providing the process learning graph and the document output to a cognitive agent

computer readable program code that accesses, by the cognitive agent and responsive to receiving a request by a user corresponding to the process, the knowledge base for the process learning graph and the document output; and

computer readable program code that assists, using the cognitive agent, the user with at least one step of the process by traversing the process learning graph and confirming process flow steps.

11. A computer program product for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the computer program product comprising:

a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising;

computer readable program code that obtains a recording of a process, wherein the recording comprises a demonstration of executing the process;

computer readable program code that generates, using the recording, the process learning graph, wherein the process learning graph identifies a process flow;

computer readable program code that generates, using the recording, the document output, wherein the document output comprises process screen transitions and process steps;

computer readable program code that expands a knowledge base of the cognitive agent by providing the process learning graph and the document output to a cognitive agent

computer readable program code that accesses, by the cognitive agent and responsive to receiving a request by a user corresponding to the process, the knowledge base for the process learning graph and the document output; and

computer readable program code that assists, using the cognitive agent, the user with at least one step of the process by traversing the process learning graph and confirming process flow steps.

12. The computer program product of claim 11, wherein the process learning graph defines user input requirements.

13. The computer program product of claim 12, wherein the process learning graph identifies format constraints for the user input requirements.

14. The computer program product of claim 11, wherein the recording comprises a script.

15. The computer program product of claim 11, wherein the recording comprises a video recording.

16. The computer program product of claim 15, wherein the video recording comprises annotations identifying at least one of:

user inputs, screen headings, and actions.

17. The computer program product of claim 16, comprising computer readable program code that extracts the process flow, the process screen transitions, and the process steps using the annotations.

18. The computer program product of claim 11, wherein the document output comprises instructions for steps to be completed on a process screen.

19. The computer program product of claim 11, comprising computer readable program code that identifies system responses based upon identified user inputs and wherein the document output comprises the system responses.

20. A method for generating a process learning graph and a document output from a recorded process for training a cognitive agent, the method comprising:

utilizing at least one processor to execute computer code that performs the steps of;

obtaining a recording of a process, wherein the recording comprises a demonstration of executing the process;

identifying at least one system response, occurring during execution of the process, based upon at least one user input;

producing, from the recording, at least one screen capture of a process screen transition;

identifying, from the recording, process screen headings;

defining, from the recording, user input requirements during execution of the process;

generating the process learning graph, wherein the process learning graph identifies a process flow comprising a plurality of process flow steps;

the plurality of process flow steps comprising the process screen headings and user input requirements;

generating the document output, wherein the document output comprises the at least one screen capture, at least one system response, and at least one process step;

providing the process learning graph and the document output to the cognitive agent;

receiving, at the cognitive agent, a message from a user, wherein the user is executing the process;

mapping the message to at least one process flow step in the process learning graph;

confirming, with the user, the mapped at least one process flow step; and

verifying, with the user, that the user performed process flow steps, contained within the process learning graph, preceding the at least one process flow step.