Voice interface for a search engine

• US 7,366,668 B1
• Filed: 02/03/2006
• Issued: 04/29/2008
• Est. Priority Date: 02/07/2001
• Status: Expired due to Term

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Milch, Brian Christopher, Brin, Sergey, Henzinger, Monika H., Franz, Alexander Mark
• Primary Examiner(s): Lerner; Martin
• Application Number:
  US11/346,182
• Time in Patent Office: 816 Days
• Field of Search: 704/231, 704/243, 704/244, 704/251, 704/252, 704/255, 704/257, 704/270, 704/236, 704/241, 704/275, 707/3, 707/4, 707/5
• US Class Current: 704/252
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/085 Methods for reducing search complexity, pruning
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A system provides search results from a voice search query. The system receives a voice search query from a user, derives one or more recognition hypotheses, each being associated with a weight, from the voice search query, and constructs a weighted boolean query using the recognition hypotheses. The system then provides the weighted boolean query to a search system and provides the results of the search system to a user.

Claims

1. A method comprising:

receiving a voice search query from a user;

deriving a plurality of recognition hypotheses from the voice search query, each recognition hypothesis being associated with a weight;

constructing a weighted boolean query using the recognition hypotheses and the weights;

providing the weighted boolean query to a search system;

receiving results for the weighted boolean query in response to providing the weighted boolean query to the search system;

determining a quantity of the received results that are related to each recognition hypothesis of the plurality of recognition hypotheses; and

discarding recognition hypotheses of the plurality of recognition hypotheses having no results to obtain a refined weighted boolean query.

2. The method of claim 1 wherein the deriving a plurality of recognition hypotheses includes:

using one or more of a language model, phonetic dictionary, or acoustic models to derive the plurality of recognition hypotheses.

3. The method of claim 2 further comprising:

updating one or more of the language model, phonetic dictionary, or acoustic models using the voice search query.

4. The method of claim 1 further comprising:

identifying a language model based on at least one characteristic associated with the user, and

wherein the deriving a plurality of recognition hypotheses includes;

using the identified language model to derive the plurality of recognition hypotheses.

5. The method of claim 1 further comprising:

discarding, prior to constructing the weighted boolean query, those recognition hypotheses associated with a weight below a threshold value.

6. The method of claim 1 wherein the weighted boolean query is a weighted OR-query.

7. The method of claim 1 further comprising:

detecting compounds in the plurality of recognition hypotheses, and

wherein the constructing a weighted boolean query includes;

constructing the weighted boolean query using the recognition hypotheses and the detected compounds.

8. The method of claim 1 further comprising:

detecting compounds in the results of the search system;

adjusting the weighted boolean query based on the detected compounds;

providing the adjusted weighted boolean query to the search system; and

providing the new results.

9. The method of claim 1 further comprising:

providing the refined weighted boolean query to the search system to obtain refined results; and

providing the refined results.

10. The method of claim 9 wherein the providing the refined results includes:

adjusting a ranking of the refined results based on the weights.

11. The method of claim 9 wherein the providing the refined results includes:

organizing the refined results based on the weights.

12. A system comprising:

means for receiving a voice search query from a user;

means for deriving a plurality of recognition hypotheses from the voice search query;

means for associating a weight with each of the recognition hypotheses;

means for constructing a weighted boolean query using the plurality of recognition hypotheses;

means for providing the weighted boolean query to a search system;

means for receiving results in response to providing the weighted boolean query to the search system;

means for determining a number of results of the received results related to each recognition hypothesis of the plurality of recognition hypotheses;

means for discarding a recognition hypothesis of the plurality of recognition hypotheses that has no related results; and

means for deriving a refined weighted boolean query using the recognition hypotheses remaining after the recognition hypothesis has been discarded.

13. A computer-readable memory device containing instructions for controlling at least one processor to perform a method, the method comprising:

receiving a voice search query;

deriving a plurality of recognition hypotheses from the voice search query, each recognition hypothesis being associated with a weight;

constructing a weighted boolean query using the plurality of recognition hypotheses;

providing weighted boolean query to a search system;

receiving results in response to providing the weighted boolean query to the search system;

determining a number of results of the received results related to each recognition hypothesis of the plurality of recognition hypotheses;

discarding a first recognition hypothesis of the plurality of recognition hypotheses that has no related results;

deriving a refined weighted boolean query using the recognition hypotheses of the plurality of recognition hypotheses remaining after the first recognition hypothesis has been discarded; and

providing the refined weighted boolean query to the search system to obtain refined results.

14. A server comprising:

a memory configured to store instructions; and

a processor configured to execute the instructions to obtain a voice search query, derive a plurality of recognition hypotheses from the voice search query, construct a boolean query using the plurality of recognition hypotheses, provide the boolean query to a search system, receive results in response to providing the boolean query to the search system, determine a quality of the received results that are related to each recognition hypothesis of the plurality of recognition hypotheses, discard a first recognition hypothesis of the plurality of recognition hypotheses based on the determined qualities of results;

refine the boolean search query using the recognition hypotheses remaining after discarding the first recognition hypothesis; and

provide the refined boolean search query to the search system to obtain new results.