Using predictive user models for language modeling on a personal device

• US 20070239637A1
• Filed: 03/17/2006
• Published: 10/11/2007
• Est. Priority Date: 03/17/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Paek, Timothy, Chickering, David
• Granted Patent: US 7,752,152 B2
• Time in Patent Office: Days
• US Class Current: 706/20
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering

Abstract

A system and method for prediction of a user goal for command/control of a personal device (e.g., mobile phone) is provided. The system employs statistical model(s) that can predict a command based, at least in part, on past user behavior (e.g., probability distribution over a set of predicates, and, optionally arguments). Further, the system can be employed with a speech recognition component to facilitate language modeling for predicting the user goal. The system can include predictive user models (e.g., predicate model and argument model) that receive a user input (e.g., utterance) and employ statistical modeling to determine the likely command without regard to the actual content of the input (e.g., utterance). The system employs features for predicting the next user goal which can be stored in a user data store. Features can capture personal idiosyncrasies or systematic patterns of usage (e.g., device-related, time-related, predicate-related, contact-specific and/or periodic features).

Claims

1. A system for the prediction of a user goal for command/control of a personal device, comprising:

a user data store that stores information regarding a particular user'"'"'s behavior with respect to the personal device; and

,

a predictive user model that receives a user input and employs statistical modeling of the information stored in the user data store to determine a likely command/control.

2. The system of claim 1, statistical modeling of the predictive user model is performed without regard to the substantive content of the user input.

3. The system of claim 1, the predictive user model determines a probability distribution over a set of potential commands/controls.

4. The system of claim 1, the predictive user model comprising a predicate model that employs statistical modeling of the information stored in the user data store to determine the likely command/control.

5. The system of claim 1, the predictive user model comprising an argument model that employs statistical modeling and the information stored in the user data store to determine a likely argument of the command/control.

6. The system of claim 1, the information stored in the user data store comprises a plurality of features associated with the particular user'"'"'s behavior.

7. The system of claim 6, the features comprising at least one of periodic, contact-specific, predicate-related, time-related and device-related features.

8. The system of claim 1, the predictive model comprises a conditional model.

9. The system of claim 8, the conditional model is a decision tree.

10. The system of claim 1, the predictive model comprises a generative model.

11. The system of claim 10, the predictive model is a naï

ve-Bayes model.

12. The system of claim 1, the predictive model learned for a general user population.

13. The system of claim 1, at least one of a parameter and a structure of the predictive model is updated in an online manner.

14. The system of claim 13, the online update is performed according to a learning rate.

15. The system of claim 1, the personal device comprising a desktop computer system, a notebook computer system, a personal digital assistant, a media player, a mobile phone, a smart phone and a remote control.

16. A method for prediction of a user goal for command/control of a personal device, comprising:

using a predictive model, determining a probability distribution over a set of potential commands/controls; and

,

employing the probability distribution with an output of a speech recognition component to determine the user goal.

17. The method of claim 16, the probability distribution determined by the predictive model is independent of a speech input.

18. The method of claim 16, the predictive model comprises a predicate model that employs statistical modeling and stored information regarding a user'"'"'s behavior to determine a likely predicate of the user goal, the predictive model further comprises an argument model that based on the likely predicate, employs statistical modeling and stored information regarding the user'"'"'s behavior to determine a likely argument of the predicate.

19. A language model system for prediction of a user goal for command/control of a personal device, comprising:

a predictive user model that receives a user input and employs statistical modeling of the information stored in the user data store to determine a likely command/control, the predictive user model providing a probability distribution;

a speech recognition component that provides a probability distribution associated with a speech input; and

,

a language model that determines the user goal based, at least in part, upon the probability distributions provided by the predictive user model and the speech recognition component.

20. The system of claim 19, the information stored in the user data store comprises a plurality of features associated with the particular user'"'"'s behavior, the features comprising at least one of periodic, contact-specific, predicate-related, time-related and device-related features.