JP2019164753A5

JP2019164753A5 -

Info

Publication number: JP2019164753A5
Application number: JP2018215657A
Authority: JP
Filing date: 2018-11-16
Publication date: 2021-05-27
Anticipated expiration: 2038-03-20

Claims

In the game the user from the owner vehicle group includes a plurality of media is advanced by issuing a Gate form fields by selecting the medium, to infer medium which can be selected by the user from the game logs stored recordable manner System
Are stored according to a user operation, obtains the game log including the game field and the medium information about the medium comprising constitute game state including the ownership vehicle group, from the game server to said game logs are stored Game log acquisition department and
And Tensor data creating unit that creates a tensor data by creating an array data based on the game log the acquired,
A learning unit that generates a learning model by performing machine learning using the created tensor data as learning data, and a learning unit.
An inference section for inferring media bodies can be selected by the user in any of the game state using the learning model,
The system.

The array data is a matrix representing the medium information included in the game state arranged in a predetermined order.
The tensor data creation unit stores information about the presence or absence of media bodies to each element of the matrix system of claim 1.

The tensor data creation unit
The system according to claim 1 or 2, wherein new tensor data is created by creating the array data based on the game log newly acquired by the game log acquisition unit.

The game log contains information for identifying the user.
The system according to any one of claims 1 to 3, wherein the tensor data creation unit creates tensor data by storing the array data for each user on a user axis indicating the type of user.

The game is a competition game in which a plurality of users play against out before Symbol game field by selecting media body from the own vehicle group each,
The system according to any one of claims 1 to 4, wherein the tensor data creation unit creates tensor data by storing the array data on the addition count axis indicating the addition count.

The type of game comprises media bodies are different from depending on the class selected by the user,
The game log contains information for identifying a class.
The system according to any one of claims 1 to 5, wherein the tensor data creation unit creates tensor data by storing the array data for each class on a class axis indicating the type of class.

The inference unit includes a user when, in yet any of the game state using the transformation matrix indicating a relationship between the medium body provided in the a being added medium body games new types of media bodies is added to the game There inferring media bodies to be selected, the system according to any one of claims 1 6.

The system according to any one of claims 1 to 7, wherein the learning model is a neural network model.

The invention according to any one of claims 1 to 8, further comprising an execution unit for executing the game program of the game using the medium inferred by the inference unit as the selection of the medium in the game state when the game progresses. Described system.

The inference unit compares the likelihood of the media body for the user to select a predetermined threshold, and outputs a media object having a more plausible likelihood,
The execution unit executes the game program by using the medium output by the inference unit as a selection of media bodies by a user, the system of claim 9.

The execution unit of the media bodies that are output by the inference unit as a selection of the media body by the user, executes the game program by using the order from medium material having a more plausible likelihood, according to claim 10 System.

The execution unit is configured to execute a game program in headless mode, the system according to any one of claims 9 11.

The owned medium group is composed of a first medium group that can be selected by the user and a second medium group that cannot be selected by the user, which is determined according to the progress of the game.
The inference unit A system according to any of inferring medium body the user selects from the first vehicle group in the game state, any one of claims 1 to 12 using the learning model.

In the game the user from the owner vehicle group includes a plurality of media is advanced by issuing a Gate form fields by selecting the medium, to infer medium which can be selected by the user from the game logs stored recordable manner Is the method of
Is created in response to user operation, obtains the game log including the game field and the medium information about the medium comprising constitute game state including the ownership vehicle group, from the game server to said game logs are stored Steps to do and
And creating a tensor data by creating an array data based on the game log the acquired,
A step of generating a learning model by performing machine learning using the created tensor data as training data, and
A step of inferring media bodies can be selected by the user in any of the game state using the learning model,
The method.

A program that causes a computer to perform each step of the method according to claim 14.