AMPER is a general Digital Me algorithm to manage the individual's state in real-time. We verify our proposed algorithm structure for the user-centric aim of improving the user's English score.
The Digital Me algorithm utilizes user state data, according to user-centric objective A (Aim), to establish M (Measure) to measure the current state of the user. It only uses data to predict the user's future states through P (Predict). After evaluating the user's future states through E (Evaluate), it is possible to maximize the user's state improvement by providing an R (Recommendation) of behavior for achieving the target state.
First, download the EdNet KT1 dataset and place the dataset in '/data/EdNet-KT1/KT1'. After that, proceed to download and run the code as follows.
git clone https://github.com/youngchan-k/AMPER.git
python main.py
Please refer to the paper and slide if you want to see more details