Crowdsourced, Grounded Language for Intent Modeling in Conversational Interfaces

• US 20130262114A1
• Filed: 04/03/2012
• Published: 10/03/2013
• Est. Priority Date: 04/03/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Brockett, Christopher John, Dolan, William Brennan, Ju, Yun-Cheng, Pantel, Patrick, Choudhury, Piali, Sophy, Noelle Mallory, Volkova, Svitlana
• Granted Patent: US 9,754,585 B2
• Time in Patent Office: Days
• US Class Current: 704/243
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

Different advantageous embodiments provide a crowdsourcing method for modeling user intent in conversational interfaces. One or more stimuli are presented to a plurality of describers. One or more sets of describer data are captured from the plurality of describers using a data collection mechanism. The one or more sets of describer data are processed to generate one or more models. Each of the one or more models is associated with a specific stimulus from the one or more stimuli.

Claims

1. A system comprising:

one or more processors;

a pre-processing component implemented on the one or more processors and configured to process data captured from a plurality of describers to generate a model; and

a conversational interface configured to be trained using the model generated by the pre-processing component.

2. The system of claim 1, wherein the model is a set of metadata documents corresponding to a stimulus.

3. The system of claim 1, wherein the conversational interface is further configured to be trained using a training algorithm in conjunction with the model.

4. The system of claim 1, wherein the conversational interface receives a verbal input that triggers a machine state change.

5. An apparatus, comprising:

one or more stimuli configured to be presented to a plurality of describers;

a data collection mechanism configured to capture describer data from the plurality of describers; and

a model generation mechanism configured to process the describer data and generate a set of metadata documents.

6. The apparatus of claim 5, wherein the one or more stimuli is a representation of a state or change of state.

7. The apparatus of claim 5, wherein the plurality of describers are a crowd of people.

8. The apparatus of claim 5, wherein the describer data is a collection of natural language descriptions from the plurality of describers, and wherein the describer data corresponds to the one or more stimuli presented to the plurality of describers.

9. The apparatus of claim 5, wherein the model generation mechanism includes one or more ranking algorithms used to process the describer data.

10. The apparatus of claim 5 further comprising:

a knowledge repository; and

a knowledge harvesting mechanism configured to interact with the knowledge repository to generate a set of word associations.

11. The apparatus of claim 10, wherein the model generation mechanism uses the set of word associations generated by the knowledge harvesting mechanism along with the one or more ranking algorithms to process the describer data and generate the set of metadata documents.

12. The apparatus of claim 5 further comprising:

a clarification mechanism configured to identify ambiguities in the set of metadata documents and generate a set of clarifying questions, wherein the set of clarifying questions are presented to the plurality of describers, and wherein the data collection mechanism captures clarifying data from the plurality of describers in response to the presentation of the set of clarifying questions.

13. The apparatus of claim 12, wherein the model generation mechanism generates a model using the set of metadata documents and the clarifying data.

14. The apparatus of claim 13, wherein the model is associated with a stimulus from the one or more stimuli.

15. A method comprising:

presenting one or more stimuli to a plurality of describers for modeling user intent in conversational interfaces;

capturing one or more sets of describer data from the plurality of describers using a data collection mechanism;

processing the one or more sets of describer data to generate one or more models, wherein each of the one or more models is associated with a specific stimulus from the one or more stimuli.

16. The method of claim 15, wherein each of the one or more sets of describer data comprises a collection of natural language descriptions associated with the one or more stimuli.

17. The method of claim 15, wherein processing the one or more sets of describer data comprises filtering and clustering the data to generate one or more sets of metadata documents, wherein each set of metadata documents corresponds to a stimulus from the one or more stimuli, and wherein each metadata document in the each set of metadata documents corresponds to a describer from the plurality of describers.

18. The method of claim 15, wherein processing the one or more sets of describer data comprises identifying ambiguities in the describer data, generating clarifying questions for the plurality of describers, capturing clarifying data form the plurality of describers in response to the clarifying questions, and generating the one or more models using the describer data and the clarifying data.

19. The method of claim 15, wherein processing the one or more sets of describer data comprises refining the describer data using a set of word associations generated by a knowledge harvesting mechanism to generate one or more sets of metadata documents.

20. The method of claim 15 further comprising:

training a conversational interface using the one or more models.