FairJob

该数据集是一个注重算法公平性的求职推荐广告数据集，旨在推动有关算法公平性的研究。该数据集记录了求职广告推荐在性别偏见方面的表现，包含20个分类特征和39个数值特征，以及一个二进制的点击标签和用户性别的代理变量。它捕捉了在一场针对性的求职活动期间，用户在5个月内的互动情况。该数据集规模为1,072,226行，任务是对求职推荐进行评估，并在点击预测中考察偏见。

This dataset is a job recommendation advertisement dataset that prioritizes algorithmic fairness, designed to advance research on algorithm fairness. It records the performance of job advertisement recommendations regarding gender bias. The dataset contains 20 categorical features, 39 numerical features, a binary click label, and a proxy variable for user gender. It captures user interactions over a 5-month period during a targeted job recruitment campaign. With a total of 1,072,226 rows, the dataset is intended for the task of evaluating job recommendations and investigating bias in click prediction.