Location-based conversational understanding

• US 10,049,667 B2
• Filed: 01/07/2016
• Issued: 08/14/2018
• Est. Priority Date: 03/31/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Heck, Larry Paul, Chinthakunta, Madhusudan, Mitby, David, Stifelman, Lisa
• Primary Examiner(s): Abebe, Daniel
• Application Number:
  US14/989,974
• Publication Number:
  US 20160118046A1
• Time in Patent Office: 950 Days
• Field of Search: 704235
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/24522 Translation of natural language queries to structured queries
• CPC Class Codes: G06F 16/24575 using context
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Location-based conversational understanding may be provided. Upon receiving a query from a user, an environmental context associated with the query may be generated. The query may be interpreted according to the environmental context. The interpreted query may be executed and at least one result associated with the query may be provided to the user.

Claims

1. A system comprising:

at least one processor; and

memory coupled to the at least one processor, the memory comprising computer executable instructions that, when executed by the at least one processor, performs a method for providing location-based conversational understanding, the method comprising;

receiving, by a computing device, a speech-based query from a user at a location;

determining whether an environmental context is associated with the location;

when it is determined that no environmental context is associated with the location;

identifying at least an acoustic interference in the speech-based query;

identifying a subject of the speech-based query;

based at least on the identified subject of the speech-based query, creating the environmental context, wherein the environmental context suppresses the acoustic interference; and

converting the speech-based query to a text-based query using the environmental context.

2. The system of claim 1, wherein the environmental context is associated with at least one of:

an understanding model to facilitate speech-to-text conversion; and

a semantic model to facilitate query execution.

3. The system of claim 1, wherein the location is determined using at least one of:

global positioning system (GPS) coordinates;

an area code associated with the user;

a zip code associated with the user; and

a proximity to a landmark.

4. The system of claim 1, wherein identifying at least an acoustic interference comprises analyzing audio of the query and identifying background noise in the audio.

5. The system of claim 1, wherein identifying the subject of the speech-based query comprises:

asking for clarification of the speech-based query from the user; and

correlating a plurality of queries, wherein the plurality of queries are identified as requesting similar information.

6. The system of claim 1, wherein creating the environmental context comprises:

associating the identified acoustic interference and the identified subject with the location; and

storing the identified acoustic interference, the identified subject and an the location information in a context database.

7. The system of claim 1, wherein converting the speech-based query comprises applying a filter for removing the acoustic interference associated with the environmental context.

8. The system of claim 1, wherein converting the speech-based query includes utilizing a Hidden Markov Model algorithm comprising statistical weightings for at least one of:

words and semantic concepts.

9. The system of claim 1, the method further comprising:

executing the text-based query according to the environmental context within a search domain associated with the identified subject; and

providing a result of the executed text-based query to the user.

10. A method for providing location-based conversational understanding, the method comprising:

receiving, by a computing device, a first speech-based query from a user at a location;

determining whether an environmental context is associated with the location;

when it is determined that no environmental context is associated with the location;

identifying at least a first acoustic interference in the first speech-based query;

identifying a subject of the first speech-based query;

based at least on the identified subject of the first speech-based query, creating a first environmental context, wherein the environmental context suppresses the acoustic interference; and

converting the first speech-based query to a text-based query using the first environmental context.

11. The method of claim 10, wherein converting the first speech-based query includes utilizing a Hidden Markov Model algorithm comprising statistical weightings for at least one of:

words likely to be associated with an understanding model; and

semantic concepts associated with a semantic model.

12. The method of claim 11, further comprising increasing the statistical weightings of one or more predicted words according to one or more previous queries received at the location.

13. The method of claim 10, wherein the first environmental context comprises an acoustic model associated with the location, and wherein the first speech-based query is adapted according to the first acoustic interference using the acoustic model.

14. The method of claim 13, wherein adapting the first speech-based query comprises:

identifying at least one background sound from the first acoustic interference;

adapting the first speech-based query to ignore the at least one background sound; and

storing the at least one background sound.

15. The method of claim 14, further comprising:

receiving a second speech-based query associated with the location;

applying the acoustic model associated with the location to the second speech-based query; and

adapting the second speech-based query to ignore the stored at least one background sound.

16. The method of claim 15, further comprising:

identifying a second acoustic interference in the second speech-based query;

based on the second acoustic interference, updating the acoustic model associated with the location; and

adapting the second speech-based query to ignore one or more background sounds in the second acoustic interference.

17. The method of claim 10, further comprising:

receiving a second speech-based query associated with the location;

creating a second environmental context based on the second speech-based query;

aggregating the first environmental context and the second environmental context into an aggregated environmental context, wherein the aggregated environmental context is associated with the locations; and

storing the aggregated environmental context.

18. The method of claim 17, wherein the aggregated environmental context comprises the subject of the first speech-based query and a subject of the second speech-based query.

19. The method of claim 18, wherein the subject of the first speech-based query is used to improve a result for the second speech-based query.

20. A computer-readable storage device storing computer executable instructions that when executed cause a computing system to perform a method for providing location-based conversational understanding, the method comprising:

receiving, by a computing device, a speech-based query from a user at a location;

determining whether an environmental context is associated with the location;

when it is determined that no environmental context is associated with the location;

identifying at least an acoustic interference in the speech-based query;

identifying a subject of the speech-based query;

based at least on the identified subject of the speech-based query, creating an environmental context, wherein the environmental context suppresses the acoustic interference; and

converting the speech-based query to a text-based query using the environmental context.