Extension to the expert conversation builder

• US 9,471,872 B2
• Filed: 06/29/2012
• Issued: 10/18/2016
• Est. Priority Date: 06/29/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Anand, Rangachari, Lee, Juhnyoung
• Primary Examiner(s): Hill, Stanley K
• Assistant Examiner(s): Traktovenko, Ilya
• Application Number:
  US13/538,634
• Publication Number:
  US 20140006319A1
• Time in Patent Office: 1,572 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

Building a runtime dialog is provided. An initial answer to an initial question is received from a user. A state of a conversation thread is created based on the initial answer. A next question to ask is automatically selected by a processor based on the updated state of the conversation thread. The next question is presented to the user. A next answer to the next question is received. The state of the conversation thread is updated based on the next answer. The selecting, presenting, receiving, and updating are repeated until a criterion is satisfied.

Claims

1. A method of building a runtime dialog, comprising:

receiving an initial answer to an initial question from a user;

creating a state of a conversation thread based on the initial answer;

selecting, automatically by a processor, a next question to ask based on the state of the conversation thread from a dialog repository and a question selection policy;

presenting the next question to the user;

receiving a next answer to the next question;

updating the state of the conversation thread based on the next answer;

repeating the selecting, presenting, receiving, and updating until a criterion is satisfied; and

wherein the question selection policy automatically resolves which question to ask next, responsive to finding that multiple questions are eligible to be asked at a point in a dialog, the question selection policy is specified by a dialog author and comprises one or more of a question ordering algorithm and question number algorithm which automatically resolves which question to ask next, a node of the dialog comprising an attached executable computer instruction that invokes a web service,

wherein the state of the conversation thread comprises a pathway through a dialog graph of a dialog knowledgebase, created as a node of the dialog graph becomes activated and represents a live dialog performed with a user, wherein multiple pathways are created representing multiple live dialog threads in a dialog with the user,

wherein the question selection policy in resolving which question to ask next from the multiple pathways kept with the user, switches between different question ordering algorithms on the multiple live dialog threads.

2. The method of claim 1, wherein the question ordering algorithm comprises depth first, breadth wise, usage based weighting, supervised machine learning, or custom scripting, or combinations thereof.

3. The method of claim 1, wherein the question number algorithm comprises fixed number, supervised machine learning, or custom scripting, or combinations thereof.

4. The method of claim 1, further comprising:

authoring a dialog and storing the dialog in the dialog repository, wherein the authoring comprises one or more of;

document-based authoring that extracts one or more nodes and links from an electronic document to create the dialog;

real-time collaborative authoring that allows two or more authors to work on the dialog in the dialog repository in real-time by merging and synchronizing the two or more authors'"'"' work;

smart device-based authoring that allows to use a gesture to create a node of the dialog;

topic-based authoring that allows an author to create a topic hierarchy containing multiple topics of a subject matter in a structured format and assign to one or more other authors for creating additional dialogs under the topics;

or

crowd-source authoring that allows the author to invite one or more other authors to join in creating the dialog;

or

combinations thereof.

5. The method of claim 1, wherein the criterion is reaching a solution to provide to the user.

6. The method of claim 1, wherein the criterion is a user stopped answering the next question.

7. The method of claim 1, further including providing a recommendation for the answer based on one or more answers received from one or more friend users of a social networking site associated with the user.

8. The method of claim 1, further including providing a recommendation for the answer based on one or more messages on an instant messaging system.

9. The method of claim 1, further including enabling the user to follow an author of a dialog in the dialog repository.

10. The method of claim 1, further including notifying the user in response to detecting a change in the dialog repository.

11. The method of claim 1, further including:

synchronizing a local dialog repository and utilization data in a local machine with a corresponding data downloaded from a master instance of a dialog manager tool on a remote server;

executing a local instance of the dialog manager tool in the local machine, wherein said steps of receiving an initial answer, selecting, presenting, receiving a next answer, updating, and repeating are performed via the local instance of the dialog manager tool and using the local dialog repository and utilization data on the local machine, wherein the local instance need not be connected to a master instance of the dialog manager tool;

uploading one or more usage statistics generated from said executing the local instance that performs said steps of receiving an initial answer, selecting, presenting, receiving a next answer, updating, and repeating, in response to detecting that the local instance is connected to the master instance in a server; and

synchronizing the local instance of the dialog manager tool and associated local dialog repository with the master instance in the server.

12. A non-transitory computer readable storage medium storing a program of instructions executable by a machine to perform a method of building a runtime dialog, comprising:

receiving an initial answer to an initial question from a user;

creating a state of a conversation thread based on the initial answer;

selecting, automatically by a processor, a next question to ask from a dialog repository based on the updated state of the conversation thread and a question selection policy;

presenting the next question to the user;

receiving a next answer to the next question;

updating the state of the conversation thread based on the next answer;

repeating the selecting, presenting, receiving, and updating until a criterion is satisfied; and

wherein the question selection policy automatically resolves which question to ask next, responsive to finding that multiple questions are eligible to be asked at a point in a dialog, the question selection policy is specified by a dialog author and comprises one or more of a question ordering algorithm and question number algorithm which automatically resolves which question to ask next, a node of the dialog comprising an attached executable computer instruction that invokes a web service,

wherein the state of the conversation thread comprises a pathway through a dialog graph of a dialog knowledgebase, created as a node of the dialog graph becomes activated and represents a live dialog performed with a user, wherein multiple pathways are created representing multiple live dialog threads in a dialog with the user, wherein the question selection policy in resolving which question to ask next from the multiple pathways kept with the user, switches between different question ordering algorithms on the multiple live dialog threads.

13. The computer readable storage medium of claim 12, wherein the question ordering algorithm comprises depth first, breadth wise, usage based weighting, supervised machine learning, or custom scripting, or combinations thereof.

14. The computer readable storage medium of claim 13, further comprising:

authoring a dialog and storing the dialog in the dialog repository, wherein the authoring comprises one or more of;

document-based authoring that extracts one or more nodes and links from an electronic document to create the dialog;

real-time collaborative authoring that allows two or more authors to work on the dialog in the dialog repository in real-time by merging and synchronizing the two or more authors'"'"' work;

smart device-based authoring that allows to use a gesture to create a node of the dialog;

topic-based authoring that allows an author to create a topic hierarchy containing multiple topics of a subject matter in a structured format and assign to one or more other authors for creating additional dialogs under the topics;

or

crowd-source authoring that allows the author to invite one or more other authors to join in creating the dialog;

or

combinations thereof.

15. The computer readable storage medium of claim 12, wherein the question selection policy comprises the question number algorithm.

16. The computer readable storage medium of claim 15, wherein the question number algorithm comprises fixed number, supervised machine learning, or custom scripting, or combinations thereof.

17. The computer readable storage medium of claim 12, wherein the criterion is reaching a solution to provide to the user.

18. The computer readable storage medium of claim 12, further including providing a recommendation for the answer based on one or more answers received from one or more friend users of a social networking site associated with the user.

19. The computer readable storage medium of claim 12, further including providing a recommendation for the answer based on one or more messages on an instant messaging system.

20. The computer readable storage medium of claim 12, further including enabling the user to follow an author of a dialog in the dialog repository.

21. The computer readable storage medium of claim 12, further including notifying the user in response to detecting a change in the dialog repository.

22. The computer readable storage medium of claim 12, further including:

matching the conversation state thread to locate one or more other users who have encountered a same conversation thread;

notifying one or more users with information associated with the same conversation thread;

enabling utilizing of one or more instant messaging systems while performing said steps of receiving an initial answer, selecting, presenting, receiving a next answer, updating, and repeating;

accessing one or more dialog-related discussion forums on a web directly;

sending a notification in response to detecting a change associated with one or more topics and nodes associated with the one or more topics in the dialog repository;

enabling the user to follow one or more other users of a dialog manager tool;

or

logging usage associated with said utilizing of one or more instance messaging systems, accessing one or more dialog-related discussion forums, and the user following one or more other users;

or

combinations thereof.

23. A system of building a runtime dialog, comprising:

a processor;

a dialog portal operable to execute on the processor and further operable to provide an interface to pose a plurality of questions to a user and receive a corresponding plurality of answers from the user;

a thread maintenance module operable to maintain a conversation state based on the received answers;

a question selection module operable to select a next question to pose based on the conversation state, a question selection policy and a dialog repository; and

wherein the question selection policy automatically resolves which question to ask next, responsive to finding that multiple questions are eligible to be asked at a point in a dialog, the question selection policy is specified by a dialog author and comprises one or more of a question ordering algorithm and question number algorithm which automatically resolves which question to ask next, a node of the dialog comprising an attached executable computer instruction that invokes a web service,

wherein the conversation comprises a pathway through a dialog graph of a dialog knowledgebase, created as a node of the dialog graph becomes activated and represents a live dialog performed with a user, wherein multiple pathways are created representing multiple live dialog threads in a dialog with the user, wherein the question selection module in resolving which question to ask next from the multiple pathways kept with the user, switches between different question ordering algorithms on the multiple live dialog threads.

24. The system of claim 23, wherein the question selection policy specifies the question ordering algorithm comprising depth first, breadth wise, usage based weighting, supervised machine learning, or custom scripting, or combinations thereof.