SDP@US1.0 - Dataset for Student Dropout Prediction

Dataset UnitelmaSapienza - Version 1.0 (May 2021)

Description: This is the dataset from the article "Hidden Space Deep Sequential Risk Prediction on Student Trajectories" by Bardh Prenkaj, Damiano Distante, Stefano Faralli and Paola Velardi (under review)

Disclaimer: Due to privacy protection regulations, all data are aggregated and fully anonymized.

Dataset Description:
The zip contains 13 folders, each folder contains the timeseries (timeseries.csv) of a specific learning course from the University of Rome Unitelma Sapienza.
Each row of a timeseries.csv file corresponds to a student. Each column is a two dimensional numpy array, which corresponds to 365 days of interactions of the student with the e-learning platform.The last column corresponds to the probability of drop-out (see the above indicated paper for more information on how this probability is thresholded to obtain a binary ground truth).
This dataset can be directly used as input for the code we released at http://iim.di.uniroma1.it/projects/hsdsrpst/
When using this dataset in your work please cite the paper it is associated to.

Authors:Bardh Prenkaj - Sapienza University of Rome, ItalyDamiano Distante - University of Rome Unitelma Sapienza, ItalyStefano Faralli - University of Rome Unitelma Sapienza, ItalyPaola Velardi - Sapienza University of Rome, Italy

License:Attribution 4.0 International (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/deed.en
For any question regarding the dataset and the source code please contact:Bardh Prenkaj ([e-mail])