Systems and methods for generating and using shared natural language libraries

Patent metadata

Field	Value
Publication number	US-12561289-B2
Application number	US-202318206567-A
Country	US
Kind code	B2
Filing date	Jun 6, 2023
Priority date	Jul 29, 2010
Publication date	Feb 24, 2026
Grant date	Feb 24, 2026

Abstract

Systems and methods for searching databases by sound data input are provided herein. A service provider may have a need to make their database(s) searchable through search technology. However, the service provider may not have the resources to implement such search technology. The search technology may allow for search queries using sound data input. The technology described herein provides a solution addressing the service provider's need, by giving a search technology that furnishes search results in a fast, accurate manner. In further embodiments, systems and methods to monetize those search results are also described herein.

First claim

What is claimed is:

1 . A method for converting speech to text, the method comprising:
receiving a sound input, by a computer system, from a client device; determining at least one natural language query from the sound input; analyzing at least a portion of the natural language query, prior to fully converting the sound input into text, to determine a topic of the natural language query; selecting a customized language model specifically trained as a large language model on the topic; and fully converting the sound input of the natural language query into text using the customized language model; wherein the customized language model is selected from a plurality of natural language libraries or sub-libraries, each aggregated from multiple service providers and stored in the cloud.

2 . The method according to claim 1 , wherein analyzing at least a portion of the natural language query to determine a topic of the natural language query comprises matching the at least a portion of the natural language query to a natural language library that is associated with the topic.

3 . The method according to claim 1 , the method further comprising:
updating content of the customized language model responsive to encountering a new type of search query.

4 . The method according to claim 3 , wherein updating the content of the customized language model is performed by a natural language library update.

5 . The method according to claim 1 , further comprising:
utilizing a natural language query processor to narrow down the possibilities in the language model to a smaller set.

6 . A non-transitory computer-readable storage medium storing instructions for converting speech to text, the instructions when executed by a computer processor performing actions comprising:
receiving a sound input, by a computer system, from a client device; determining at least one natural language query from the sound input; analyzing at least a portion of the natural language query, prior to fully converting the sound input into text, to determine a topic of the natural language query; selecting a customized language model specifically trained as a large language model on the topic; and fully converting the sound input of the natural language query into text using the customized language model; wherein the customized language model is selected from a plurality of natural language libraries or sub-libraries, each aggregated from multiple service providers and stored in the cloud.

7 . The non-transitory computer-readable storage medium according to claim 6 , wherein analyzing at least a portion of the natural language query to determine a topic of the natural language query comprises matching the at least a portion of the natural language query to a natural language library that is associated with the topic.

8 . The non-transitory computer-readable storage medium according to claim 6 , the actions further comprising:
updating content of the customized language model responsive to encountering a new type of search query.

9 . The non-transitory computer-readable storage medium according to claim 8 , wherein updating the content of the customized language model is performed by a natural language library update.

10 . The non-transitory computer-readable storage medium according to claim 6 , the actions further comprising:
utilizing a natural language query processor to narrow down the possibilities in the language model to a smaller set.

Assignees

• Soundhound Inc
• Soundhound Ai Ip Llc

Inventors

• Mohajer Keyvan

Classifications

• G10L2015/0635

  updating or merging of old and new templates; Mean values; Weighting · CPC title

• G10L15/183

  using context dependencies, e.g. language models · CPC title

• G10L15/063

  Training · CPC title

• G06F16/3344 Primary

  using natural language analysis · CPC title

• G06F16/3334

  Selection or weighting of terms from queries, including natural language queries · CPC title