System and method for determining query intent

• US 9,373,322 B2
• Filed: 04/10/2013
• Issued: 06/21/2016
• Est. Priority Date: 04/10/2013
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Goussard, Jacques-Olivier
• Primary Examiner(s): Shah, Paras D
• Application Number:
  US13/859,983
• Publication Number:
  US 20140309993A1
• Time in Patent Office: 1,168 Days
• Field of Search: 704/9, 704/270.1, 704/270, 704/275, 379/88.01, 707/705, 707/713, 707/718, 707/771, 707/769, 707/776
• US Class Current: 1/1
• CPC Class Codes: G06F 16/33 Querying
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30
• CPC Class Codes: G10L 15/063 Training

Abstract

A method for training a system is provided. The method may include storing one or more backend communication logs, each of the one or more backend communication logs including a user query and a corresponding backend query. The method may further include parsing the one or more backend communication logs to extract statistical information and generating a mapping between each user query and a corresponding set of language tags. The method may also include sorting the one or more backend communication logs based upon, at least in part, the extracted statistical information.

Claims

1. A computer-implemented method for use with an interactive voice response system comprising:

storing, using one or more computing devices implementing an interactive voice response application, a plurality of backend communication logs, each of the plurality backend communication logs including a user query and a corresponding backend query;

parsing, using the one or more computing devices, the plurality of backend communication logs to extract statistical information;

translating the extracted statistical information into a natural language tag;

generating, using the one or more computing devices, a mapping between each user query and a corresponding set of language tags based on the translating;

sorting, using the one or more computing devices, the plurality of backend communication logs based upon, at least in part, the extracted statistical information and the mapping;

receiving, using the one or more computing devices, a current user query, at the interactive voice response application, and the corresponding backend query;

updating the interactive voice response application by dynamically updating the mapping, wherein dynamically updating the mapping is based upon, at least in part, a current user query being received and a determined commonality between the current user query and at least one of a previous user query, wherein the determined commonality associates the current user query with the corresponding set of language tags for the previous user query;

determining a query intent for a subsequent user query, received via the interactive voice response application, based on the sorted backend communication logs and the updated mapping; and

providing, via the interactive voice response application, a database response to the user based upon, at least in part, the determined query intent and the backend communication log.

2. The method of claim 1, wherein the extracted statistical information includes at least one of high runner queries and data priors.

3. The method of claim 1 wherein the extracted statistical information includes at least one of mandatory fields, optional fields, and disambiguation templates.

4. The method of claim 3, wherein the extracted statistical information includes a disambiguation template configured to group one or more user queries and their corresponding backend queries, wherein the one or more user queries represent a user dialog.

5. The method of claim 4 further comprising:

extracting interaction templates from the dialog, wherein the interaction templates include tailored dialog strategies.

6. The method of claim 3 wherein the extracted statistical information includes a disambiguation template configured to group a plurality of user queries and their corresponding backend queries belonging to the same task context.

7. The method of claim 6 further comprising:

extracting interaction templates from the groups of the plurality of user queries and their corresponding backend queries.

8. The method of claim 1 wherein parsing the one or more backend communication logs includes parsing an SQL communication log.

9. The method of claim 1, wherein parsing the one or more backend communication logs includes parsing a REST communication log.

10. A non-transitory computer readable storage medium having stored thereon instructions that when executed by a processor perform one or more operations for use with an interactive voice response system, the operations comprising:

storing, using one or more computing devices implementing an interactive voice response application, a plurality of backend communication logs, each of the plurality backend communication logs including a user query and a corresponding backend query;

parsing, using the one or more computing devices, the plurality of backend communication logs to extract statistical information;

translating the extracted statistical information into a natural language tag;

generating, using the one or more computing devices, a mapping between each user query and a corresponding set of language tags based on the translating;

sorting, using the one or more computing devices, the plurality of backend communication logs based upon, at least in part, the extracted statistical information and the mapping;

receiving, using the one or more computing devices, a current user query, at the interactive voice response application, and the corresponding backend query;

updating the interactive voice response application by dynamically updating the mapping, wherein dynamically updating the mapping is based upon, at least in part, a current user query being received and a determined commonality between the current user query and at least one of a previous user query, wherein the determined commonality associates the current user query with the corresponding set of language tags for the previous user query;

determining a query intent for a subsequent user query, received via the interactive voice response application, based on the sorted backend communication logs and the updated mapping; and

providing, via the interactive voice response application, a database response to the user based upon, at least in part, the determined query intent and the backend communication log.

11. The non-transitory computer readable storage medium of claim 10, wherein the extracted statistical information includes at least one of high runner queries and data priors.

12. The non-transitory computer readable storage medium of claim 10 wherein the extracted statistical information includes at least one of mandatory fields, optional fields, and disambiguation templates.

13. The non-transitory computer readable storage medium of claim 12 wherein the extracted statistical information includes a disambiguation template configured to group a plurality of user queries and their corresponding backend queries belonging to the same task context.

14. The non-transitory computer readable storage medium of claim 13 wherein operations further comprise:

extracting interaction templates from the groups of the plurality of user queries and their corresponding backend queries.

15. The non-transitory computer readable storage medium of claim 10 wherein parsing the one or more backend communication logs includes parsing an SQL communication log.

16. The non-transitory computer readable storage medium of claim 10, wherein parsing the one or more backend communication logs includes parsing a REST communication log.

17. An interactive voice response system comprising:

one or more processors configured to execute one or more operations, the operations including;

storing, using one or more computing devices implementing an interactive voice response application comprising at least a portion of an interactive voice response system, one or more a plurality of backend communication logs, each of the one or more plurality backend communication logs including a user query and a corresponding backend query;

parsing, using the one or more computing devices, the one or more plurality of backend communication logs to extract statistical information;

translating the extracted statistical information into a natural language tag;

generating, using the one or more computing devices, a mapping between each user query and a corresponding set of language tags based on the translating;

sorting, using the one or more computing devices, the one or more plurality of backend communication logs based upon, at least in part, the extracted statistical information and the mapping;

receiving, using the one or more computing devices, a current user query, at the interactive voice response application, and the corresponding backend query; and

updating the interactive voice response system application by dynamically updating the mapping, wherein dynamically updating the mapping is based upon, at least in part, a current user query being received and a determined commonality between the current user query and at least one of a previous user query, wherein the determined commonality associates the current user query with the corresponding set of language tags for the previous user query;

determining a query intent for a subsequent user query, received via the interactive voice response application, based on the sorted backend communication logs and the updated mapping; and

providing, via the interactive voice response application, a database response to the user based upon, at least in part, the determined query intent and the backend communication log.