Determining match type for query tokens

Patent metadata

Field	Value
Publication number	US-9830353-B1
Application number	US-201313778312-A
Country	US
Kind code	B1
Filing date	Feb 27, 2013
Priority date	Feb 27, 2013
Publication date	Nov 28, 2017
Grant date	Nov 28, 2017

Abstract

Example processes for determining a match type include: receiving a query token from a content provider; consulting a database stored in memory to determine past performance of the query token for associated content, where the past performance includes performance information indicating how well the query token performed for different match types, where a match type indicates a way that components of the query token match components of another query token in order to achieve a token match; determining, based on the performance information, differences in performance between ones of the different match types; comparing the differences in performance to a threshold; and selecting a match type for the query token based on the comparing.

First claim

What is claimed is:

1. A method performed by one or more processing devices, comprising:
receiving, from a content provider, a query token that controls distribution of content provided by the content provider;
consulting a database stored in memory to determine past performance of the query token when used to distribute content, the past performance comprising performance information indicating how well the query token performed for different match types, where a match type defines what is required to constitute a token match between components of the query token and components of a search query that is submitted by a user;
identifying, by the one or more processing devices, a specificity value of the query token that quantifies how specific the query token is based on how rare the query token is in a particular corpus of words;
determining, by the one or more processing devices and based on the performance information, differences in performance of the query token between the different match types;
determining that the differences in performance meet a threshold difference in performance;
in response to the determination that the differences in performance meet the threshold difference in performance, comparing the identified specificity value of the query token that quantifies how specific the query token is to a threshold specificity;
selecting, based on the comparing, a match type for the query token, the selecting including:
selecting an exact match type for the query token when the specificity of the query token is less than the threshold specificity, wherein selection of the exact match type requires all components of a user submitted query to match the query token for the content to be eligible for distribution; and
selecting a phrase match type or broad match type for the query token when the specificity of the query token is greater than the threshold specificity, wherein each of the phrase match type and the broad match type require less of a match than the exact match type;

assigning, by the one or more processing devices, the match type to the query token, where assigning comprises storing the match type in association with the query token and content associated with the query token; and
comparing the query token to received input of a search query from a user different from the content provider in accordance with the match type to identify the content for output to a display device.

2. The method of claim 1 , wherein selecting the match type comprises:
selecting a match type having a difference in performance that exceeds the threshold difference in performance; and
in a case that more than one match type exceeds the threshold difference in performance, selecting a match type among the more than one match type that provides most match options for the query token.

3. The method of claim 1 , wherein selecting the match type comprises:
determining that none of the differences in performance exceed the threshold difference in performance;
comparing performance information for each of the different match types to a second threshold difference in performance different from the threshold difference in performance; and
selecting a match type for the query token based on the comparing to the second threshold difference in performance.

4. The method of claim 3 , wherein selecting a match type for the query token based on the comparing to the second threshold difference in performance comprises:
selecting a match type having performance information that exceeds the second threshold difference in performance; and
in a case that more than one match type exceeds the threshold difference in performance, selecting a match type among the more than one match type that provides most match options for the query token.

5. The method of claim 1 , wherein the performance information comprises at least one content metric, the at least one content metric comprising at least one of click-through rate for content associated with the query token, a cost-per-click associated with the content, or a cost-per-acquisition associated with the content.

6. The method of claim 1 , wherein the different match types comprise at least exact match, phrase match, and broad match, where phrase match requires a sequence of components in the query token to match a sequence of components of the other token, and where broad match requires components of the query token to match components of the other token;
wherein determining, based on the performance information, differences in performance between ones of the different match types comprises determining, for the query token, a first difference between performance information for exact match and performance information for phrase match, and determining a second difference between performance information for exact match and broad match; and
wherein both the first difference and the second difference are compared to the threshold difference in performance.

7. The method of claim 1 , wherein the query token comprises a keyword or phrase associated with online advertising, the selected match type being associated with the keyword and the online advertising.

8. The method of claim 1 , wherein the query token comprises at least one of an image, audio, video, or text.

9. A method performed by one or more processing devices, comprising:
storing a database in memory comprising past performance of query tokens when used to distribute content provided in response to online searching, the past performance comprising performance information indicating how well query tokens performed for different match types, where a match type defines what is required to constitute a token match between components of the query token and components of a search query that is submitted by a user;
receiving, from a content provider, a query token that controls distribution of content provided by the content provider;
identifying, by the one or more processing devices, a specificity value of the query token that quantifies how specific the query token is based on how rare the query token is in a particular corpus of words;
determining, by the one or more processing devices and based on the performance information, differences in performance of the query token between the different match types;
determining that the differences in performance meet a threshold difference in performance;
assigning, by the one or more processing devices, a match type to the received query token in response to the determination that the differences in performance meet the threshold difference in performance, and comparing a specificity of the query token to a threshold specificity, where assigning comprises storing the match type in association with the query token and content associated with the query token, and where assignment of an exact match type requires all components of a user submitted query to match the query token for the content to be eligible for distribution, wherein assignment of each of a phrase match type and a broad match type require less of a match than the exact match type; and
comparing the query token to received input of a search query from a user different from the content provider in accordance with the match type to identify the content for output to a display device.

10. The method of claim 9 , further comprising:
identifying one or more query tokens in the database that have a relationship to the received token; and
retrieving, from the database, the selected past performance information for the one or more query tokens.

11. One or more machine-readable storage devices storing instructions that are executable by one or more processing devices to perform operations comprisin

Assignees

• Dalmia Aditya
• Kwatra Nipun
• Tiwari Pranav Kumar
• Panesar Kiran S
• Google Inc

Inventors

• Dalmia Aditya
• Kwatra Nipun
• Tiwari Pranav Kumar
• Panesar Kiran S

Classifications

• G06Q30/0242 Primary

  Determining effectiveness of advertisements · CPC title

• G06F16/2453 Primary

  Query optimisation · CPC title

• G06F17/30442 Primary

  Physics · mapped topic

• G06F16/951

  Indexing; Web crawling techniques · CPC title