Building topic-oriented audiences

Abstract

Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium for creating and querying for audience data. A method includes: receiving data associated with each firing of a pixel as a result of a display of a publisher resource, wherein the data includes an identifier for the publisher, a URL associated with the resource, and a user identifier associated with a user device on which the resource was displayed; determining for each pixel firing one or more topics associated with a given resource; storing the user identifier in association with the determined one or more topics in an audience data structure; receiving a query to identify a segment of the audience; identifying one or more topics in the audience data structure based at least in part on terms of the query; and identifying user identifiers that are in an audience segment defined by the query.

1 . (canceled) 
     
     
         2 . A method performed by one or more servers comprising:
 providing executable code that causes a client device to evaluate content presented by a resource when the resource is presented by the client device, wherein the resource is provided by a given publisher;   collecting, from each particular client device among multiple different client devices that visit the resource, data extracted from the resource by the particular client device during the evaluation that occurs in response to execution of the executable code when the resource is presented by the particular client device, wherein the data extracted specifies different content that was presented to the client devices over time;   determining, for each execution of the executable code, topics of the different content presented on the resource over time based on the data specifying content that was presented on the resource that was extracted by the execution of the executable code;   receiving a query from the publisher to identify a segment of an audience that has visited one or more pages of the publisher's website that are related to a given topic;   in response to receiving the query:
 identifying one or more matching topics in a data structure that are relevant to the given topic specified by the query; 
 identifying, in the data structure, identifiers corresponding to a set of client devices that presented content from the matching topics when the resource was presented by the set of client devices; and 
 providing a given content item to the set of client devices that are associated with the matching topics in the data structure. 
   
     
     
         3 . The method of  claim 2  wherein at least one of the one or more resources is a web resource. 
     
     
         4 . The method of  claim 2  wherein at least one of the one or more resources is an application presented on a mobile device. 
     
     
         5 . The method of  claim 2  wherein collecting data extracted from the resource includes collecting one or more of the title or a portion of the resource. 
     
     
         6 . The method of  claim 2  wherein determining, for each execution of the executable code, topics includes:
 identifying a semantic space that includes a plurality of topic clusters, each topic cluster including one or more topics that are within a predetermined distance in the semantic space; 
 determining one or more topic clusters for the resource; and 
 storing a user identifier in association with the determined one or more topic clusters. 
 
     
     
         7 . The method of  claim 6  wherein determining the one or more topic clusters includes determining a probability that the resource is associated with one of the determined one or more topic clusters. 
     
     
         8 . The method of  claim 7  wherein storing the user identifier with the specific topic cluster includes storing the probability along with the user identifier in a given topic cluster. 
     
     
         9 . A system comprising:
 one or more processors; and   one or more memory elements including instructions that when executed cause the one or more processors to perform operations including:
 providing executable code that causes a client device to evaluate content presented by a resource when the resource is presented by the client device, wherein the resource is provided by a given publisher; 
   collecting, from each particular client device among multiple different client devices that visit the resource, data extracted from the resource by the particular client device during the evaluation that occurs in response to execution of the executable code when the resource is presented by the particular client device, wherein the data extracted specifies different content that was presented to the client devices over time;   determining, for each execution of the executable code, topics of the different content presented on the resource over time based on the data specifying content that was presented on the resource that was extracted by the execution of the executable code;   receiving a query from the publisher to identify a segment of an audience that has visited one or more pages of the publisher's website that are related to a given topic;   in response to receiving the query:
 identifying one or more matching topics in a data structure that are relevant to the given topic specified by the query; 
 identifying, in the data structure, identifiers corresponding to a set of client devices that presented content from the matching topics when the resource was presented by the set of client devices; and 
 providing a given content item to the set of client devices that are associated with the matching topics in the data structure. 
   
     
     
         10 . The system of  claim 9  wherein at least one of the one or more resources is a web resource. 
     
     
         11 . The system of  claim 9  wherein at least one of the one or more resources is an application presented on a mobile device. 
     
     
         12 . The system of  claim 9  wherein collecting data extracted from the resource includes collecting one or more of the title or a portion of the resource. 
     
     
         13 . The system of  claim 9  wherein determining, for each execution of the executable code, topics includes:
 identifying a semantic space that includes a plurality of topic clusters, each topic cluster including one or more topics that are within a predetermined distance in the semantic space; 
 determining one or more topic clusters for the resource; and 
 storing a user identifier in association with the determined one or more topic clusters. 
 
     
     
         14 . The system of  claim 13  wherein determining the one or more topic clusters includes determining a probability that the resource is associated with one of the determined one or more topic clusters. 
     
     
         15 . The system of  claim 14  wherein storing the user identifier with the specific topic cluster includes storing the probability along with the user identifier in a given topic cluster. 
     
     
         16 . A computer program product tangibly embodied in a computer-readable storage device and comprising instructions that, when executed by a processor, cause the processor to perform operations including:
 providing executable code that causes a client device to evaluate content presented by a resource when the resource is presented by the client device, wherein the resource is provided by a given publisher;   collecting, from each particular client device among multiple different client devices that visit the resource, data extracted from the resource by the particular client device during the evaluation that occurs in response to execution of the executable code when the resource is presented by the particular client device, wherein the data extracted specifies different content that was presented to the client devices over time;   determining, for each execution of the executable code, topics of the different content presented on the resource over time based on the data specifying content that was presented on the resource that was extracted by the execution of the executable code;   receiving a query from the publisher to identify a segment of an audience that has visited one or more pages of the publisher's website that are related to a given topic;   in response to receiving the query:
 identifying one or more matching topics in a data structure that are relevant to the given topic specified by the query; 
 identifying, in the data structure, identifiers corresponding to a set of client devices that presented content from the matching topics when the resource was presented by the set of client devices; and 
 providing a given content item to the set of client devices that are associated with the matching topics in the data structure. 
   
     
     
         17 . The computer program product of  claim 16  wherein at least one of the one or more resources is a web resource. 
     
     
         18 . The computer program product of  claim 16  wherein at least one of the one or more resources is an application presented on a mobile device. 
     
     
         19 . The computer program product of  claim 16  wherein collecting data extracted from the resource includes collecting one or more of the title or a portion of the resource. 
     
     
         20 . The computer program product of  claim 16  wherein determining, for each execution of the executable code, topics includes:
 identifying a semantic space that includes a plurality of topic clusters, each topic cluster including one or more topics that are within a predetermined distance in the semantic space; 
 determining one or more topic clusters for the resource; and 
 storing a user identifier in association with the determined one or more topic clusters. 
 
     
     
         21 . The computer program product of  claim 20  wherein determining the one or more topic clusters includes determining a probability that the resource is associated with one of the determined one or more topic clusters.