Assigning terms of interest to an entity

• US 8,589,399 B1
• Filed: 03/26/2012
• Issued: 11/19/2013
• Est. Priority Date: 03/25/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Lee, Jason, Stern, Tamara I., Donaker, Gregory J., Blair-Goldensohn, Sasha J.
• Primary Examiner(s): Beausoliel, Jr., Robert
• Assistant Examiner(s): HOANG, HAU HAI
• Application Number:
  US13/430,624
• Time in Patent Office: 603 Days
• Field of Search: 707E17083- 84, 707/730, 707/741, 707/737
• US Class Current: 707/737
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing

Abstract

The subject matter of this specification can be embodied in, among other things, a method that includes identifying resources relating to an entity, where each resource includes multiple terms and is included in a corpus of resources relating to multiple entities. Candidate terms from the resources for potentially associating with the entity and a category associated with the entity are identified. A relative frequency of the candidate terms in the identified resources is compared to a frequency of the candidate terms associated with other entities. Each of the candidate terms are weighted, for example, based on a source of the candidate term and the relative frequency of the candidate term. A weighted frequency of each candidate term is calculated based on the weights, and candidate terms are selected as representative terms for the entity based on the weighted frequency.

Claims

1. A computer-implemented method for determining characteristic terms for an entity, the method comprising:

identifying one or more resources relating to the entity, each resource including a plurality of terms;

identifying a plurality of candidate terms from the plurality of terms;

identifying a business type associated with the entity;

identifying one or more resources associated with other entities having the same business type;

determining, via one or more processors and for each of the plurality of candidate terms;

a first frequency with which the candidate term appears in the one or more resources related to the entity;

a second frequency with which the candidate term appears in the one or more resources associated with other entities in the identified business type;

a relative frequency for the candidate term based on the first frequency and the second frequency; and

a weighted relative frequency for the candidate term based on the relative frequency of the candidate term and how frequently the entity is selected as a search result when presented in response to a search query that includes the candidate term;

identifying, via one or more processors, one or more of the candidate terms as characteristic terms for the entity based on the weighted relative frequency of each candidate term; and

associating the identified characteristic terms with the entity in a data repository.

2. The method of claim 1 further comprising:

identifying, for at least a first candidate term, a set of related candidate terms, and determining the first and second frequencies and the weighted relative frequency for the first candidate term based on the set of related candidate terms.

3. The method of claim 2 wherein the set of related candidate terms includes a term that is at least one of:

a plural of the first candidate term;

a substantially similar semantic variation of the first candidate term;

a synonym of the first candidate term;

or

a subphrase of the first candidate term.

4. The method of claim 1 wherein identifying the plurality of candidate terms includes omitting, from the plurality of terms in the one or more resources related to the entity, a predefined set of stop words.

5. The method of claim 1 wherein identifying the plurality of candidate terms includes omitting, from the plurality of terms in the one or more resources related to the entity, terms that fall within one or more predefined classes of terms.

6. The method of claim 1 wherein identifying the plurality of candidate terms includes omitting from the plurality of terms in the one or more resources related to the entity one or more of:

terms that refer to a location of the entity;

terms that are variations of a name of the entity;

contact information associated with the entity;

terms included in a list of stop words associated with the business type associated with the entity;

terms that are common in documents associated with the determined business type;

or

temporal terms.

7. The method of claim 1 wherein identifying the business type for the entity and identifying one or more resources associated with entities having the same business type comprises determining the location of the entity and identifying one or more resources associated with entities within a particular geographical area near the location of the entity.

8. The method of claim 1 wherein identifying the business type for the entity and identifying one or more resources associated with entities having the same business type comprises determining a business category for the entity and identifying one or more resources associated with entities having a same business category as the entity.

9. The method of claim 8 further comprising:

identifying a second business category associated with the entity;

identifying one or more resources associated with other entities in the second business category;

determining, via one or more processors and for each of the plurality of candidate terms;

a third frequency with which the candidate term appears in the one or more resources associated with other entities in the second business category;

a second relative frequency for the candidate term based on the first frequency and the third frequency; and

a weighted second relative frequency for the candidate term based on the second relative frequency of the candidate term and how frequently the entity is selected as a search result when presented in response to a search query that includes the candidate term;

identifying, via one or more processors, one or more of the candidate terms as characteristic terms for the entity in the second business category based on the weighted second relative frequency of each candidate term; and

associating the identified characteristic terms with the entity and the second business category in a data repository.

10. The method of claim 1 wherein the plurality of resources include one or more of:

web documents relating to the entity;

a database storing attributes associated with the entity;

a database storing terms used to search for a web document relating to the entity;

a database storing terms identifying features of neighborhoods;

or

a database storing terms associated with the identified business type.

11. The method of claim 1 wherein identifying, via one or more processors, one or more of the candidate terms as characteristic terms for the entity based on the weighted relative frequency of each candidate term, further comprises:

filtering candidate terms with a weighted relative frequency that is less than a threshold.

12. The method of claim 1 wherein identifying the plurality of candidate terms further includes identifying terms in the one or more resources relating to the entity that appear in a predetermined list of terms associated with the determined business type.

13. The method of claim 1 further comprising one or more of:

using the characteristic terms for the entity to generate search results for search queries that include the characteristic terms;

displaying one or more of the characteristic terms in association with displaying data identifying the entity;

displaying suggested refinements to a search query received from a user;

generating a set of characteristic terms for a geographical area that includes the entity from the characteristic terms for the entity;

soliciting user feedback on the entity based on the characteristic terms;

or

grouping the entity with other entities that share one or more characteristic terms.

14. A computer storage medium encoded with a computer program, the program comprising instructions that when executed by data processing apparatus cause the data processing apparatus to perform operations including:

identifying a set of candidate terms for potentially associating with an entity, wherein the set of candidate terms are included within one or more resources relating to the entity;

identifying at least one of an entity type category for the entity or an entity location of the entity;

determining a first frequency with which each of the candidate terms appears in the one or more resources relating to the entity;

determining a second frequency with which each of the candidate terms appears in one or more resources relating to the identified entity type category or the identified entity location;

determining a relative frequency for each of the candidate terms based on the first frequency and the second frequency;

determining a weighted relative frequency for each of the candidate terms based on the relative frequency of each candidate term and how frequently the entity is selected as a search result when presented in response to a respective search query that includes each candidate term;

selecting one or more candidate terms as being characteristic terms for the entity based at least in part on the determined weighted relative frequency of each candidate term; and

storing data identifying the selected characteristic terms for the entity as being associated with the entity.

15. The computer storage medium of claim 14 wherein the instructions to cause the data processing apparatus to select one or more candidate terms as characteristic terms for the entity further comprise instructions to cause the data processing apparatus to eliminate one or more candidate terms having a weighted relative frequency that falls below a threshold.

16. A system comprising:

one or more memories storing instructions; and

one or more processors configured to execute the instructions stored in the one or more memories in order to;

identify a set of candidate terms for potentially associating with an entity, wherein the set of candidate terms are included within one or more resources that include information about the entity; and

identify at least one of an entity type category for the entity or an entity location of the entity;

determine a first frequency with which each of the candidate terms appears in the one or more resources relating to the entity;

determine a second frequency with which each of the candidate terms appears in one or more resources relating to the identified entity type category or the identified entity location;

determine a relative frequency for each of the candidate terms based on the first frequency and the second frequency;

determine a weighted relative frequency for each of the candidate terms based on the relative frequency of each candidate term and how frequently the entity is selected as a search result when presented in response to a respective search query that includes each candidate term;

select one or more candidate terms as being characteristic terms for the entity based on the weighted relative frequency of each candidate term; and

store data associating the characteristic terms with information about the entity in a data repository.

17. The system of claim 16 wherein the one or more processors are configured to select one or more candidate terms as characteristic terms for the entity by eliminating one or more candidate terms having a weighted relative frequency that is less than a threshold.

18. The system of claim 16 wherein the one or more processors are further configured to eliminate from the set of candidate terms at least one of:

candidate terms describing a location of the entity;

candidate terms relating to a name of the entity;

or

candidate terms in a predefined list of terms associated with at least one of the entity type category or the entity location.