SYSTEMS AND METHODS FOR PROVIDING IMPROVED SEARCH FUNCTIONALITY ON A CLIENT DEVICE

• US 20160357830A1
• Filed: 09/29/2015
• Published: 12/08/2016
• Est. Priority Date: 06/05/2015
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Salim, David, Hornkvist, John, Prakash, Vipul Ved
• Granted Patent: US 10,360,902 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24532 of parallel queries
• CPC Class Codes: G06F 16/338 Presentation of query results
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology

Abstract

A system, method, and machine readable medium are described for providing improved search functionality on a client device.

Claims

1. A machine implemented method for providing a search functionality on a computer, the method comprising:

receiving a search query, wherein the search query is a user'"'"'s voice input;

transmitting, over one or more networks, the search query to one or more data processing systems to produce one or more speech to text recognitions and execute one or more parallel search queries based on the one or more speech to text recognitions, the parallel search queries being used to search one or more remotely stored indices;

receiving, from the one or more data processing systems, the one or more speech to text recognitions and results of the one or more parallel search queries;

determining one or more search queries based on the speech to text recognitions of the received query;

executing the one or more search queries to search a locally stored index using the one or more speech to text recognitions; and

presenting combined results of the one or more search queries and the parallel search queries.

2. The machine implemented method of claim 1, wherein the presentation of the combined results of the one or more search queries and the parallel search queries is based on the speech to text recognition that has the highest probability of being accurate.

3. The machine implemented method of claim 2, wherein the probability that a speech to text recognition is accurate is based on comparing the results of the one or more search queries, the parallel search queries, and the respective speech to text recognitions on which they are based.

4. The machine implemented method of claim 3, wherein a higher probability of accurate recognition is determined when a speech to text recognition produces more results for its corresponding search queries than other speech to text recognitions.

5. The machine implemented method of claim 2, wherein the probability of accurate speech to text recognition is determined by resolving ambiguities in the user'"'"'s voice input.

6. The machine implemented method of claim 5, further comprising:

wherein resolving ambiguities includes distinguishing between likely utterances and unlikely utterances in the user'"'"'s voice input; and

determining whether an unlikely utterance produces more local search results than a likely utterance.

7. The machine implemented method of claim 5, further comprising:

wherein resolving ambiguities includes comparing voice input with historical data from previous searches by the user.

8. The method of claim 1 wherein the one or more remotely stored indices include a set of one or more information domains including:

(a) social media postings;

(b) current news;

(c) weather;

(d) maps;

(e) sports scores;

(f) stock information;

or (g) current movies.

9. The method of claim 8 wherein the locally stored index includes one or both of content and metadata of a plurality of different files, having different file types, created by a plurality of different applications.

10. A non-transitory machine readable medium storing executable program instructions which when executed by a data processing system cause the system to perform a method for providing a search functionality, the method comprising:

receiving a search query, wherein the search query is a user'"'"'s voice input;

transmitting, over one or more networks, the search query to one or more data processing systems to produce one or more speech to text recognitions and execute one or more parallel search queries based on the one or more speech to text recognitions, the parallel search queries being used to search one or more remotely stored indices;

receiving, from the one or more data processing systems, the one or more speech to text recognitions and results of the one or more parallel search queries;

determining one or more search queries based on the speech to text recognitions of the received query;

executing the one or more search queries to search a locally stored index using the one or more speech to text recognitions; and

presenting combined results of the one or more search queries and the parallel search queries.

11. The medium of claim 10, wherein the presentation of the combined results of the one or more search queries and the parallel search queries is based on the speech to text recognition that has the highest probability of being accurate.

12. The medium of claim 11, wherein the probability that a speech to text recognition is accurate is based on comparing the results of the one or more search queries, the parallel search queries, and the respective speech to text recognitions on which they are based.

13. The medium of claim 12, wherein a higher probability of accurate recognition is determined when a speech to text recognition produces more results for its corresponding search queries than other speech to text recognitions.

14. The medium of claim 11, wherein the probability of accurate speech to text recognition is determined by resolving ambiguities in the user'"'"'s voice input.

15. The medium of claim 14, the method further comprising:

wherein resolving ambiguities includes distinguishing between likely utterances and unlikely utterances in the user'"'"'s voice input; and

determining whether an unlikely utterance produces more local search results than a likely utterance.

16. The medium of claim 14, further comprising:

wherein resolving ambiguities includes comparing voice input with historical data from previous searches by the user.

17. The medium of claim 10 wherein the one or more remotely stored indices include a set of one or more information domains including:

(a) social media postings;

(b) current news;

(c) weather;

(d) maps;

(e) sports scores;

(f) stock information;

or (g) current movies.

18. The medium of claim 17 wherein the locally stored index includes one or both of content and metadata of a plurality of different files created by a plurality of different applications and being of different file types.

19. A data processing system comprising:

means for receiving a search query, wherein the search query is a user'"'"'s voice input;

means for transmitting, over one or more networks, the search query to one or more data processing systems to produce one or more speech to text recognitions and execute one or more parallel search queries based on the one or more speech to text recognitions, the parallel search queries being used to search one or more remotely stored indices;

means for receiving, from the one or more data processing systems, the one or more speech to text recognitions and results of the one or more parallel search queries;

means for determining one or more search queries based on the speech to text recognitions of the received query;

means for executing the one or more search queries to search a locally stored index using the one or more speech to text recognitions; and

means for presenting combined results of the one or more search queries and the parallel search queries.

20. The system of claim 19, wherein the presentation of the combined results of the one or more search queries and the parallel search queries is based on the speech to text recognition that has the highest probability of being accurate;

and wherein the probability that a speech to text recognition is accurate is based on comparing the results of the one or more search queries, the parallel search queries, and the respective speech to text recognitions on which they are based.

21. The system of claim 20, wherein a higher probability of accurate recognition is determined when a speech to text recognition produces more results for its corresponding search queries than other speech to text recognitions.