Back to results
US 11,455,660 B2Granted
Extraction device, extraction method, and non-transitory computer readable storage medium
AssigneeYahoo Japan Corporation
InventorsTakashi Tashiro, Shumpei Okura
Life of this patent
Estimating the term…
Abstract
The extraction device according to the present application includes an acquisition unit and an extraction unit. The acquisition unit acquires behavior histories of users being candidates to whom content is to be distributed. The extraction unit extracts target users expected to perform specified behavior, based on a behavior history designated by a content provider, of the behavior histories acquired by the acquisition unit. For example, the extraction unit uses a model for determining the degree of similarity between a user performing the specified behavior and a target user expected to perform specified behavior based on the behavior history designated by the content provider in order to extract the target users.
Claims7
A system comprising:
- a web server;
- a plurality of user terminals associated with a plurality of users, the plurality of user terminals being connected to the web server via a network, each of the user terminals transmitting a cookie to the web server over the network when accessing the web server to acquire a web page, the cookie including a user identifier and a behavior history of the associated user;
- and
- an extraction device connected via the network to the web server, the extraction device comprising:
- a memory;
- and
- a processor operatively coupled with the memory, the processor being programmed to:
- acquire, over the network, the user identifiers and behavior histories of the plurality of users from the web server;
- store the user identifiers and behavior histories of the plurality of users in the memory;
- provide a user interface to a content provider terminal over the network, the user interface being configured to prompt a content provider to input designations of a specified behavior and an explanatory behavior;
- receive, over the network from the content provider terminal, the specified behavior and the explanatory behavior designated by the content provider;
- generate a model by identifying similarities between behavior histories including the specified behavior, and using an association between the explanatory behavior and the specified behavior, the model enabling prediction that one of the plurality of users will perform the specified behavior based on its behavior history;
- determine, for each of the users, whether a similarity between the user's behavior history and the model is not less than a predetermined threshold;
- and
- in response to the similarity of one or more users being not less than the predetermined threshold, extract the one or more users as the candidates predicted to perform the specified behavior and to whom content is to be distributed.
12 more elements
An extraction method performed by a system, comprising:
- transmitting to a web server, over a network, a cookie from each of a plurality of user terminals associated with a plurality of users when the user terminal acquires a web page from the web server, the cookie including a user identifier and a behavior history of the associated user;
- and
- using a computer:
- acquiring, over the network, the user identifiers and behavior histories of the plurality of users from the web server;
- storing the user identifiers and behavior histories of the plurality of users in a memory;
- providing a user interface to a content provider terminal over the network, the user interface being configured to prompt a content provider to input designations of a specified behavior and an explanatory behavior;
- receiving, over the network from the content provider terminal, the specified behavior and the explanatory behavior designated by the content provider;
- generating a model by identifying similarities between behavior histories including the specified behavior, and using an association between the explanatory behavior and the specified behavior, the model enabling prediction that one of the plurality of users will perform the specified behavior based on its behavior history;
- determining, for each of the users, whether a similarity between the user's behavior history and the model is not less than a predetermined threshold;
- and
- in response to the similarity of one or more users being not less than the predetermined threshold, extract the one or more users as candidates predicted to perform the specified behavior and to whom content is to be distributed.
8 more elements
A system comprising:
- a web server;
- a plurality of user terminals of users connected to the web server via a network, each of the user terminals transmitting over the network a cookie to the web server when accessing the web server to acquire a web page, the cookie including a user identifier (ID) and a behavior history of the corresponding user;
- and
- a non-transitory computer readable storage medium having stored therein an extraction program, the extraction program causing a computer to perform:
- acquiring, over the network, user IDs and behavior histories of users from the web server;
- accumulating the acquired behavior histories of the users in a memory, based on user IDs as information for identification of the users;
- providing a user interface to a content provider terminal over the network, the user interface being configured to prompt a content provider to input designations of a specified behavior and an explanatory behavior;
- receiving, over the network from the content provider terminal, the specified behavior and the explanatory behavior designated by the content provider;
- generating a model by identifying similarities between behavior histories including the specified behavior, and using an association between the explanatory behavior and the specified behavior, the model enabling prediction that users will perform the specified behavior based on behavior history;
- determining, for each of the users, whether a similarity between the user's behavior history and the model is not less than a predetermined threshold;
- and
- in response to the similarity of one or more users being not less than the predetermined threshold, extract the one or more users as candidates predicted to perform the specified behavior and to whom content is to be distributed.
9 more elements
Dependencies are inferred from the claim text.