Interactive reformulation of speech queries

• US 10,176,219 B2
• Filed: 03/13/2015
• Issued: 01/08/2019
• Est. Priority Date: 03/13/2015
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Dumoulin, Benoit, Ahmadi, Ali, Parthasarathy, Sarangarajan, Craswell, Nick, Ozertem, Umut, Shokouhi, Milad, Raghunathan, Karthik, Jones, Rosie
• Primary Examiner(s): Spieler, William
• Application Number:
  US14/657,568
• Publication Number:
  US 20160267128A1
• Time in Patent Office: 1,397 Days
• Field of Search: None
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 25/48 specially adapted for particular use

Abstract

Methods and systems are provided for providing alternative query suggestions. For example, a spoken natural language expression may be received and converted to a textual query by a speech recognition component. The spoken natural language expression may include one or more words, terms, and/or phrases. A phonetically confusable segment of the textual query may be identified by a classifier component. The classifier component may determine at least one alternative query based on identifying at least the phonetically confusable segment of the textual query. The classifier may further determine whether to suggest the at least one alternative query based on whether the at least one alternative query is sensical and/or useful. When it is determined to suggest the at least one alternative query, the at least one alternative query may be provided to and displayed on a user interface display.

Claims

1. A system comprising:

at least one processor; and

memory encoding computer executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method for providing alternative query suggestions, the method comprising;

receiving a spoken natural language expression;

converting the spoken natural language expression to a textual query;

identifying at least a phonetically confusable segment of the textual query, wherein identifying the phonetically confusable segment of the textual query comprises determining a context of the textual query, wherein the context is associated with a topic of the textual query;

identifying at least one alternative query, wherein the at least one alternative query is homophonically related to the phonetically confusable segment of the textual query;

determining whether the at least one alternative query makes logical sense based on the context by evaluating the at least one alternative query against a context store by measuring engagement of the at least one alternative query; and

upon determining the at least one alternative query makes logical sense, providing the at least one alternative query to a user interface display.

2. The system of claim 1, wherein identifying at least the phonetically confusable segment of the textual query comprises using at least one of metaphones, historical data, and a plurality of speech recognition confidence scores.

3. The system of claim 1, further comprising determining whether a speech recognition confidence score associated with a best guess textual query candidate meets a threshold.

4. The system of claim 3, wherein when it is determined that the speech recognition confidence score meets the threshold, the method further comprising returning search results associated with the best guess textual query candidate.

5. The system of claim 1, wherein determining whether to suggest the at least one alternative query comprises determining whether the at least one alternative query is useful.

6. The system of claim 1, wherein determining whether to suggest the at least one alternative query comprises identifying co-occurrences of the textual query and the at least one alternative query in a dictionary store.

7. The system of claim 1, wherein determining whether to suggest the at least one alternative query comprises determining whether the at least one alternative query provides search results that are different than search results provided by the textual query.

8. The system of claim 1, wherein the at least one alternative query includes at least one of a plurality of textual query candidates.

9. The system of claim 1, wherein the at least one alternative query includes replacing the phonetically confusable segment of the textual query with a phonetically-similar term.

10. The system of claim 1, wherein the at least one alternative query includes replacing the phonetically confusable segment of the textual query with at least one feature identified in at least one of a context store and a dictionary store.

11. A system comprising:

at least one processor; and

memory encoding computer executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method for interactive reformulation of speech queries, the method comprising;

presenting a query and one or more search results associated with the query on a user interface display of a client device;

receiving a selection of a segment of the presented query on the user interface display;

receiving, at the client device, a voice suggestion for replacing the selected segment of the presented query, wherein the voice suggestion and the selected segment are recorded as an associated pair; and

displaying one or more alternative terms corresponding to the voice suggestion on the user interface display of the client device, wherein the one or more alternative terms make logical sense when evaluated against a context of the presented query, wherein evaluating the context of the presented query comprises comparing the one or more alternative terms to historical data in a context store by measuring engagement of one or more of alternative terms.

12. The system of claim 11, further comprising:

receiving a selection of at least one of the one or more alternative terms displayed on the user interface display of the client device;

replacing, at the client device, the selected segment with the selected at least one alternative term to provide a reformulated query;

receiving one or more search results associated with the reformulated query; and

displaying the reformulated query and the one or more search results associated with the reformulated query on the user interface display of the client device.

13. The system of claim 11, further comprising:

sending the voice suggestion to a speech recognition component; and

receiving, at the client device, a reformulated query and one or more search results associated with the reformulated query, wherein the reformulated query includes the presented query with the selected segment of the presented query replaced with one or more alternative terms corresponding to the voice suggestion.

14. The system of claim 13, wherein the one or more alternative terms corresponding to the voice suggestion are identified based on metaphones, historical data, and a plurality of speech recognition confidence scores.

15. The system of claim 11, wherein the presented query is a best guess textual query from a speech recognition component.

16. The system of claim 11, wherein the presented query is an alternative query suggested by a classifier component.

17. A method for interactive reformulation of speech queries including at least one processor, and memory encoding computer executable instructions that, when executed by at least one processor, cause the at least one processor to perform the method, the method comprising:

presenting a query and one or more search results associated with the query on a user interface display of a client device;

receiving a selection of at least a segment of the presented query on the user interface display;

receiving, at the client device, a voice suggestion for replacing the selected segment of the presented query, wherein the voice suggestion and the selected segment are recorded as an associated pair;

sending the voice suggestion to a speech recognition component;

receiving, at the client device, a reformulated query and one or more search results associated with the reformulated query, wherein the reformulated query includes the presented query with the selected segment of the presented query replaced with a textual representation of the voice suggestion, and wherein the reformulated query makes logical sense when evaluated against a topical context of the presented query by measuring engagement of the reformulated query; and

displaying the reformulated query and the one or more search results associated with the reformulated query on the user interface display of the client device.

18. The method of claim 17, further comprising receiving one or more alternative query suggestions to the reformulated query.

19. The method of claim 17, further comprising:

receiving, at the client device, one or more alternative terms for replacing the selected segment of the presented query; and

displaying the one or more alternative terms on the user interface display of the client device.