Business listing search

• US 8,831,930 B2
• Filed: 10/27/2010
• Issued: 09/09/2014
• Est. Priority Date: 10/13/2006
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Strope, Brian, Byrne, William J., Beaufays, Francoise
• Primary Examiner(s): MCFADDEN, SUSAN IRIS
• Application Number:
  US12/913,151
• Publication Number:
  US 20110047139A1
• Time in Patent Office: 1,413 Days
• Field of Search: 704/10
• US Class Current: 704/10
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/197 Probabilistic grammars, e.g. word n-grams
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A method of operating a voice-enabled business directory search system includes receiving category-business pairs, each category-business pair including a business category and a specific business, and establishing a data structure having nodes based on the category-business pairs. Each node of the data structure is associated with one or more business categories and a speech recognition language model for recognizing specific businesses associated with the one or more businesses categories.

Claims

1. A method comprising:

collecting, by one or more processors, information about associations of specific businesses with categories from keyword searches;

establishing, by the one or more processors, speech recognition language models based on the information, each language model being associated with one or more categories, each language model for recognizing specific businesses associated with the one or more categories; and

recognizing, by the one or more processors, specific businesses in a speech utterances using the language models.

2. The method of claim 1, further comprising establishing, by the one or more processors, a hierarchical tree having nodes, each node being associated with one or more of the categories and one of the speech recognition language models.

3. The method of claim 1 wherein the keyword searches comprise at least one of web searches, intranet searches, and desktop searches.

4. A method comprising:

receiving, by one or more processors, a speech input having information about a business category and an identifier of a specific business;

mapping, by the one or more processors, the type of business in the speech input to nodes in a data structure, each node being associated with one or more business categories and a speech recognition language model; and

recognizing, by the one or more processors, the identifier of the specific business using one or more language models determined based on the mapping.

5. The method of claim 4 wherein the mapping comprises, for each of some of the nodes, determining a similarity score representing a similarity between the business category in the speech input and the one or more business categories associated with the node.

6. The method of claim 5, further comprising generating, by the one or more processors, weights for the language models based on the similarity scores.

7. The method of claim 4, further comprising finding, by the one or more processors, a particular node having a highest similarity to the business category in the speech input, and using a first language model associated with the particular node and a second language model associated with a parent node of the particular node to recognize the identifier.

8. An apparatus comprising:

a voice-enabled user interface to receive a speech input having information about a business category and an identifier of a specific business;

a mapping module to compare the business category to a plurality of nodes of a data structure, each node being associated with one or more business categories and a speech recognition language model; and

a speech recognition module to recognize the identifier of the specific business using one or more language models determined based on the mapping.

9. The apparatus of claim 8 wherein the mapping module determines, for each of some of the nodes, a similarity score between the business category in the speech input and the one or more business categories associated with the node.

10. The apparatus of claim 8 wherein the mapping module generates weights for the one or more language models based on the similarity scores.

11. The apparatus of claim 8 wherein the mapping module finds a particular node having a highest similarity to the business category in the speech input, and uses a first language model associated with the particular node and a second language model associated with a parent node of the particular node to recognize the identifier.

12. An apparatus comprising:

means for mapping information about a business category to a plurality of nodes of a hierarchical tree and generating weight values for the nodes, each node being associated with one or more business categories and a language model for recognizing specific businesses associated with the one or more business categories; and

a speech recognition engine to recognize a specific business in a speech input using one or more language models determined based on the mapping.

13. The apparatus of claim 12 wherein the mapping means determines weight values for the nodes based on the mapping, and the one or more language models are weighted by the weight values.