Initializing a Workspace for Building a Natural Language Understanding System

• US 20150032441A1
• Filed: 07/26/2013
• Published: 01/29/2015
• Est. Priority Date: 07/26/2013
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Marcus, Jeffrey N.
• Granted Patent: US 10,229,106 B2
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars

Abstract

Designing a natural language understanding (NLU) model for an application from scratch can be difficult for non-experts. A system can simplify the design process by providing an interface allowing a designer to input example usage sentences and build an NLU model based on presented matches to those example sentences. In one embodiment, a method for initializing a workspace for building an NLU system includes parsing a sample sentence to select at least one candidate stub grammar from among multiple candidate stub grammars. The method can include presenting, to a user, respective representations of the candidate stub grammars selected by the parsing of the sample sentence. The method can include enabling the user to choose one of the respective representations of the candidate stub grammars. The method can include adding to the workspace a stub grammar corresponding to the representation of the candidate stub grammar chosen by the user.

Claims

1. A method for initializing a workspace for building a Natural Language Understanding (NLU) system, the method comprising:

parsing a sample sentence to select at least one candidate stub grammar from among multiple candidate stub grammars;

presenting, to a user, respective representations of the at least one candidate stub grammar selected by the parsing of the sample sentence;

enabling the user to choose one of the respective representation of the candidate stub grammars; and

adding to the workspace a stub grammar corresponding to the representation of the candidate stub grammar chosen by the user.

2. The method of claim 1, wherein the respective representations are names of the candidate stub grammars.

3. The method of claim 1, further comprising presenting each respective representation of the candidate stub grammar to the user by visually presenting each of the respective representations of the candidate stub grammars relative to a corresponding part of the sample sentence.

4. The method of claim 1, wherein parsing the sample sentence includes parsing the sample sentence to generate a set of annotations including mentions.

5. The method of claim 4, wherein the annotations further include mentions and intents.

6. The method of claim 1, wherein the presented respective representations have a semantic match to the parsed sample sentence above a threshold.

7. The method of claim 1, wherein, if the at least one candidate grammar is one candidate grammar, the method further comprises:

skipping presenting respective representations of the one candidate grammar; and

skipping enabling the user to choose one of the respective representations;

wherein adding to the workplace a stub grammar includes adding to the workplace the one candidate grammar.

8. A system for initializing a workspace for building a Natural Language Understanding (NLU) system, the system comprising:

a parsing module configured to parse a sample sentence to select at least one candidate stub grammar from among multiple candidate stub grammars;

a presentation module configured to present, to a user, respective representations of the at least one candidate stub grammar selected by the parsing of the sample sentence;

a user interface module configured to enable the user to choose one of the respective representation of the candidate stub grammars; and

a workspace building module configured to add a chosen grammar, chosen by the user from among the respective presented representations, to the workspace.

9. The system of claim 8, wherein the respective representations are names of the candidate stub grammars.

10. The system of claim 8, wherein the presentation module is further configured to visually present each of the respective representations of the candidate stub grammars relative to a corresponding part of the sample sentence.

11. The system of claim 8, wherein the parsing module is further configured to parse the sample sentence to generate a set of annotations including mentions.

12. The system of claim 11, wherein the annotations further include mentions and intents.

13. The system of claim 8, wherein the presented respective representations have a semantic match to the parsed sample sentence above a threshold.

14. The system of claim 8, if the at least one candidate grammar is one candidate grammar, the presentation module and user interface module are disabled, and the workplace building module is configured to add the one candidate grammar.

15. A non-transitory computer-readable medium configured to store instructions for initializing a workspace for building a Natural Language Understanding (NLU) system, the instructions, when loaded and executed by a processor, cause the processor to:

parse a sample sentence to select at least one candidate stub grammar from among multiple candidate stub grammars;

present, to a user, respective representations of the at least one candidate stub grammar selected by the parsing of the sample sentence;

enable the user to choose one of the respective representation of the candidate stub grammars; and

add to the workspace a stub grammar corresponding to the representation of the candidate stub grammar chosen by the user.

16. The non-transitory computer-readable medium of claim 15, wherein the respective representations are names of the candidate stub grammars.

17. The non-transitory computer-readable medium of claim 15, wherein the instructions further cause the processor to present each respective representation of the candidate stub grammars to the user by visually presenting each of the respective representations of the candidate stub grammars relative to a corresponding part of the sample sentence.

18. The non-transitory computer-readable medium of claim 15, wherein instructions further cause the processor to parse the sample sentence to generate a set of annotations including mentions.

19. The non-transitory computer-readable medium of claim 18, wherein the annotations further include mentions and intents.

20. The non-transitory computer-readable medium of claim 15, wherein the presented respective representations have a semantic match to the parsed sample sentence above a threshold.