Assistant system, assistant method, and non-transitory computer readable recording medium

An assistant method for providing a service in response to a request from a user by processes executed by a computer including one or more memories which stores a plurality of pieces of content data and a preference model representing preference of the user regarding the service, each of the plurality of pieces of content data linked with preference information regarding the service,

• wherein the assistant method includes executing a first process and a second process by the computer,
• wherein the first process includes:
• selecting notification content data from the plurality of pieces of content data, the notification content data being content data to be notified to the user, wherein the notification content data is selected based on having a low degree of matching with the preference of the user;
• acquiring an evaluation of the user for the notification content data;
• and
• updating the preference model based on the evaluation for the notification content data and the preference information linked with the notification content data,
• wherein the second process includes:
• extracting, based on the preference model of the user, one or more pieces of match content data having a high degree of matching with the preference of the user from the plurality of pieces of content data;
• and
• providing the service using the one or more pieces of match content data, and
• wherein the selecting the notification content data in the first process includes:
• acquiring, based on the preference model of the user, a deflection degree of the preference of the user;
• and
• selecting the notification content data by using either a normal selecting procedure or an exceptional selecting procedure, the normal selecting procedure being selecting content data at random from the plurality of pieces of content data, the exceptional selecting procedure being selecting content data from one or more pieces of unmatched content data having the low degree of matching with the preference of the user among the plurality of pieces of content data, where a frequency of using the exceptional selecting procedure increases as the deflection degree increases.

A non-transitory computer readable recording medium on which a computer program for causing a computer to execute processes for providing a service in response to a request from a user is recorded,

• wherein the computer includes one or more memories storing a plurality of pieces of content data and a preference model representing preference of the user regarding the service, each of the plurality of pieces of content data linked with preference information regarding the service, and
• wherein the computer program, when executed by the computer, causes the computer to execute a first process and a second process,
• wherein the first process includes:
• selecting notification content data from the plurality of pieces of content data, the notification content data being content data to be notified to the user, wherein the notification content data is selected based on having a low degree of matching with the preference of the user;
• acquiring an evaluation of the user for the notification content data;
• and
• updating the preference model based on the evaluation for the notification content data and the preference information linked with the notification content data,
• wherein the second process includes:
• extracting, based on the preference model of the user, one or more pieces of match content data having a high degree of matching with the preference of the user from the plurality of pieces of content data;
• and
• providing the service using the one or more pieces of match content data, and
• wherein the selecting the notification content data in the first process includes:
• acquiring, based on the preference model of the user, a deflection degree of the preference of the user;
• and
• selecting the notification content data by using either a normal selecting procedure or an exceptional selecting procedure, the normal selecting procedure being selecting content data at random from the plurality of pieces of content data, the exceptional selecting procedure being selecting content data from one or more pieces of unmatched content data having the low degree of matching with the preference of the user among the plurality of pieces of content data, where a frequency of using the exceptional selecting procedure increases as the deflection degree increases.