Published May 9, 2024 | Version 1.0.0
Dataset Open

Unfair Inequality in Education: A Benchmark for AI-Fairness Research (Aequitas WP7 Use Case S2)

Description

Unfair Inequality in Education: A Benchmark for AI-Fairness Research

This is the repository for the code and dataset of the paper intitled "Unfair Inequality in Education: A Benchmark for AI-Fairness Research" submitted to the DEMO track of the 27TH European Conference on Artificial Intelligence (ECAI).

Abstract

This paper proposes a novel benchmark specifically designed for AI fairness research in education. It can be used for challenging tasks aimed at improving students' performance and reducing dropout rates which are also discussed in the paper to emphasize significant research directions. By prioritizing fairness, this benchmark aims to foster the development of bias-free AI solutions, promoting equal educational access and outcomes for all students.

Structure

benchmark contains:

  • the proposed dataset (dataset.csv), and
  • the mask for dealing with missing values (missing_mask.csv).

raw_data includes:

  • the original dataset (original.csv), and
  • the intermediate stages of the pre-processing pipeline (split and pre_processed).

res contains the documentation, including:

  • the transformation mapping each column of the original dataset to the proposed one, along with the missingness category and original text (meta_data_mapping.csv), and
  • the value type and domains of each column of the original, intermediate-stage, and proposed datasets (rispectively, meta_data_original.json and meta_data_merged.json and meta_data_final.json).

src contains the source code for running the pre-processing and corresponding analysis:

  • pre_processing and statscontain the code for the two corresponding tasks, and
  • pre_processing.py and split.py are two entry points.

Finally, Dockerfile and requirements.txt set up the environment for running the applications across multiple platforms and with Python, respectively.

Files

students-dataset.zip

Files (94.4 MB)

Name Size Download all
md5:c1d0f63904c3360a147fb1d9e07af4b3
94.4 MB Preview Download

Additional details

Funding

European Commission
AEQUITAS – ASSESSMENT AND ENGINEERING OF EQUITABLE, UNBIASED, IMPARTIAL AND TRUSTWORTHY AI SYSTEMS 101070363

Software

Repository URL
https://github.com/aequitas-aod/ull-dataset-preprocessing
Programming language
Python
Development Status
Active