Santa leads the paradigm shift in the TOEIC education with AI technology.
We Stay Motivated
Santa motivates learners by visualizing their current learning status in contrast to their own target scores as well as comparison to average user score trends. Users receive personalized comments and reports, which encourage them to solve more questions than they normally would when studying with books.
4M+
Actual users
300M+
TOEIC Learning Data
165
Average Score Increase on 20 Hours of Learning
95%
Predicted Score Accuracy
Our approach
Result-oriented
Riiid does not just talk about our vision or daydream about success. We lead the paradigm shift through tangible results. We will be able to tell the next generation: ”We used to have workbooks”, “we used to take internet lectures.”
Santa analyzes and conducts research on the following 5 models.
- Knowledge tracing (prediction of the learner's response: predicting the percentage of correct answers to multiple-choice questions)
- Score Prediction (predicting the learner's ability: predicting real-time expected score)
- Recommender system including Global RS and Cycle RS (recommendation optimized for improving learning effect) System)
- Dropout prediction (predicting learner's departure from learning)
The model predicts the learning period based on the correlation between your predicted score and target score and recommends the shortest learning path.
Vulnerability analysis predicts the correct rate of tags based on content tags, analyzes the average, and diagnoses and recommends that the concept of the tag is weak.
The user analysis report includes the TOEIC prediction score, the five-point index consisting of listening, reading, vocabulary, grammar, and structure, the proficiency status checked by the probability of correct answers for each part, Vulnerability analysis, correct answer analysis by LC accent, RC solution time analysis, etc. are provided. Compare the average correct probability of users with the same score.
Just by conducting 12 free TOEIC tests, Riiid's TOEIC performance results are analyzed and derived in detail with 6 visual data. The results are divided into predicted scores, TOEIC skills, skills for each part, weak concepts, LC intonation analysis, and RC problem-solving time. Through this, you will be able to know in detail your weaknesses that you previously knew little about, so you can assign your efforts for the TOEIC test more efficiently.
We do not suggest a separate learning target amount for each score. Although it can be sorted by scores, there is a difference in each individual's learning tendency and weakness, so we are aiming to achieve personalized learning.
As a result of having analyzed users, studying with our solution for 20 hours increased about 165 points, regardless of the score range. This is an average number, and the scope of increase varies from person to person.
Unlike the existing learning that provides curriculum according to a set period of time, such as two-week or 30-day courses, our solution provides what is only necessary, so the learning time can be reduced as much as possible according to individual efforts.
Another benefit is that with our solution, there is no limit to how much you can learn at once, so you can coordinate your learning time.
The deep learning prediction model applied to Santa uses the solution data for at least 7 questions (view selection, time required, whether or not to watch the learning material, etc.) with high accuracy in which view to choose for more than 10,000 questions that the user has not solved. predict.
This technology is applied to all educational fields, enabling 'accurate skill diagnosis' of each individual within 10 minutes of any subject.
In addition, Santa proposes the world's first deep learning model to predict the churn rate in a mobile learning environment. Predict when learning users will stop learning and at what problems they will turn off and leave your app. If it is applied to the problem recommendation model and has the same learning efficiency, it helps users to focus on learning by recommending problems with a low churn rate.
With a single user ID, you can access on your devices such as mobile, tablet, and PC to study at home, on the bus or subway, and anywhere. In addition, there is no limit to the number of devices that can be used per user ID, so you can learn as much as you want and wherever you want. However, in order to save learning data and provide customized learning, you must have an internet connection.