Learning platform of obtaining memory retention rate using machine learning

1 . A learning platform for obtaining a memory retention rate using machine learning, comprising a memory assist server configured to:

generate a dataset of questions and answers comprising one or more question-and-answer pairs;

collect log data from a user terminal corresponding to the dataset of questions and answers, wherein the log data includes:

1) A content identification (ID) identifying a type of a content,

2) correct answers,

3) incorrect answers, and

4) a number of steps that a user has tried on the dataset;

extract a first pre-processed data by filtering the log data to only obtain data points where the number of steps is identical;

obtain a second pre-processed data by converting the correct answers and the incorrect answers in the first pre-processed data into their respective counts;

group the second pre-processed data points having the same content ID;

extract learning data including a percentage of correct answers from the grouped second pre-processed data, wherein the learning data further comprises i) an order R of learning trials of the dataset of questions and answers, and ii) an elapsed time T calculated by a difference between a completion time of an n-th learning trial (R) and a completion time of an (n+1)-th learning trial, wherein n is a positive integer;

convert the order R of learning trials and the elapsed time T in the learning data to their respective inverses;

train an artificial neural network using the learning data; and

calculate a memory retention rate of the dataset of questions and answers using the trained artificial neural network.

2 . The learning platform of claim 1 , wherein the learning data further includes a public memory retention rate, wherein the public memory retention rate is a median of percentages of correct answers about the dataset of questions and answers by a plurality of users when the content ID and the steps are all identical, wherein each of the plurality of users indicates one who has taken a test more than once.

3 . The learning platform of claim 1 , wherein the memory assist server further includes:

an input data generator generating input data about the dataset of questions and answers when the user terminal connects to the memory assist server; and

a memory calculator inputting the input data to the artificial neural network trained by a machine learning portion to calculate the memory retention rate.

4 . The learning platform of claim 3 , wherein the input data includes: i) an order R′ of a learning trial of the dataset of questions and answers and ii) an elapsed time T′ that has elapsed since a completion time of the learning trial.