Information processing method, information processing system, and information terminal to assist user learning
AssigneeSony Group Corporation
InventorsKazuki Kawamura, Junichi Rekimoto, Kojiro Kashiwa
Life of this patent
Estimating the term…
Abstract
Provided is an information processing method that includes an input step of inputting time-series media information representing a movement or conduct of a learning user, a first determination step of determining a learning level of the user on the basis of the time-series media information, and an output step of outputting a portion in the time-series media information at which the movement or action by the user is different from a reference movement or action, on the basis of the learning level of the user determined in the first determination step.
Claims17
An information processing method, comprising:
- inputting, to a trained machine learning model, time-series media information representing an action of a user,
- wherein the trained machine learning model:
- extracts, based on a self-training learning model, a feature quantity of the inputted time-series media information;
- and
- classifies, based on a supervised-training learning model, a learning level of the user based on the extracted feature quantity;
- receiving, from the trained machine learning model, the classified learning level of the user as an output of the trained machine learning model;
- and
- outputting, based on the classified learning level of the user, a portion in the time-series media information at which the action of the user is different from a reference action.
An information processing system, comprising:
- at least one processor configured to:
- receive, as an input, time-series media information that represents an action of a user;
- input the received time-series media information to a trained machine learning model,
- wherein the trained machine learning model:
- extracts, based on a self-training learning model, a feature quantity of the inputted time-series media information;
- and
- classifies, based on a supervised-training learning model, a learning level of the user based on the extracted feature quantity;
- receive, from the trained machine learning model, the classified learning level of the user as an output of the trained machine learning model;
- and
- output, based on the classified learning level of the user, a portion in the time-series media information at which the action of the user is different from a reference action.
An information terminal, comprising:
- a sensor configured to:
- detect an action of a user;
- and
- acquire time-series media information that represents the action of the user;
- a communication interface configured to:
- transmit the time-series media information to a trained machine learning model, wherein the trained machine learning model:
- extracts, based on a self-training learning model, a feature quantity of the inputted time-series media information;
- and
- classifies, based on a supervised-training learning model, a learning level of the user based on the extracted feature quantity;
- receive, from the trained machine learning model, the classified learning level of the user;
- and
- receive, from an external apparats, a portion in the time-series media information at which the action of the user is different from a reference action, wherein the portion in the time-series media information is based on the classified learning level of the user;
- and
- a display screen configured to display the received portion in the time-series media information.
A non-transitory computer-readable medium having stored thereon, computer-executable instructions which, when executed by a processor, cause the processor to execute operations, the operations comprising:
- inputting, to a trained machine learning model, time-series media information representing an action of a user,
- wherein the trained machine learning model:
- extracts, based on a self-training learning model, a feature quantity of the inputted time-series media information;
- and
- classifies, based on a supervised-training learning model, a learning level of the user based on the extracted feature quantity;
- receiving, from the trained machine learning model, the classified learning level of the user as an output of the trained machine learning model;
- and
- outputting, based on the classified learning level of the user, a portion in the time-series media information at which the action of the user is different from a reference action.
Dependencies are inferred from the claim text.