Method and architecture for consolidated database search for input recognition systems

• US 7,197,494 B2
• Filed: 04/10/2003
• Issued: 03/27/2007
• Est. Priority Date: 10/15/2002
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Wang, Kuansan, Ollason, David G., Bhatia, Siddharth
• Primary Examiner(s): Rimell; Sam
• Assistant Examiner(s): WILLOUGHBY, ALICIA M
• Application Number:
  US10/411,059
• Publication Number:
  US 20040073540A1
• Time in Patent Office: 1,447 Days
• Field of Search: 707/3, 707/10, 707/100, 704/236, 704/257
• US Class Current: 707/705
• CPC Class Codes: G06V 30/274 Syntactic or semantic context, e.g. balancing
• CPC Class Codes: G10L 15/193 Formal grammars, e.g. finite state automata, context free grammars or word networks
• CPC Class Codes: Y10S 707/959 Network
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation

Abstract

A method and system are provided for recognizing user input. User input is applied to at least one model to identify at least one decoded value found in a grammar. The decoded value and at least one database value stored in the grammar are then returned. The database value corresponds to a database value found in a database that is separate from the grammar. Under one embodiment, the database is associated with an application that accesses database values in the database. Under a further embodiment, two database values are returned with the decoded value where one of the database values corresponds to a database value in a first database and the other database value corresponds to a database value in a second database.

Claims

1. A method of recognizing input, the method comprising:

receiving values representing user input;

applying the input values to at least one model that provides likelihoods of recognition associated with possible recognition values based on the input values to identify at least one recognition value, wherein the possible recognition values that likelihoods are provided for are limited to those recognition values found in a grammar that is separate from the model, the grammar comprising a set of entries, at least one entry comprising possible recognition values and values to be output when the possible recognition values are identified from user input values, the values to be output comprising a first database value and a second database value that are different from the possible recognition values, the first database value corresponding to a database value found in a first database that is separate from the grammar and that is associated with a first application that accesses database values in the first database, and the second database value corresponding to a database value found in a second database that is separate from the grammar and that is associated with a second application that accesses database values in the second database; and

returning the identified recognition value, a first database value and a second database value from an entry in the grammar.

2. The method of claim 1 wherein applying the input values to at least one model comprises applying the input values to an acoustic model.

3. The method of claim 1 wherein returning the recognition value, the first database value and the second database value comprises returning a mark-up language page.

4. The method of claim 1 further comprising updating the grammar by accessing the first database to obtain updated database values.

5. The method of claim 4 wherein updating the grammar further comprises not changing the recognition values in the grammar.

6. A system for decoding user input, the system comprising:

a user agent providing a user interface for obtaining input from a user;

a first database having first database values;

a first application capable of accessing database values in the first database;

a second database having second database values;

a second application capable of accessing database values in the second database;

a recognition grammar, separate from the first database and the second database, comprising a set of entries, at least one entry comprising allowed decoded values that can be represented in the input from a user, at least one associated database value from the first database and at least one associated database value from the second database; and

a recognition process capable of identifying at least one decoded value in the grammar by determining likelihoods of recognition associated with decoded values in the grammar based on user input, wherein the recognition process limits the decoded values for which likelihoods are determined to decoded values in the grammar and wherein the recognition process is further capable of returning the identified decoded value, an associated database value that is from the first database and is stored in the grammar and an associated value that is from the second database and is stored in the grammar to the user agent.

7. The system of claim 6 wherein the user agent obtains input from the user based on instructions received from the first application.

8. The system of claim 6 wherein the recognition process returns the decoded value and the database values in a mark-up language page.

9. The system of claim 6 wherein the user agent is executed on a first computing device and the recognition process is executed on a second computing device.

10. The system of claim 9 wherein the first computing device communicates with the second computing device through a network.