2022:NeurIPS:Fair Ranking with Noisy Protected Attributes

[Hamedloghmani]

This issue is dedicated to the summary of a paper that I found recently(Fair Ranking with Noisy Protected Attributes)
It offers a new Fair Ranking framework and mathematically proves the guarantees mentioned in the paper. I took a more high-level approach in summarizing this time.
2022_NeurIPS_Fair Ranking With Noisy Protected Attributes.pdf

[hosseinfani]

@Hamedloghmani
Thank you for the nice summary. Do you think we can apply it in our work? For instance, in our dblp or github dataset, we don't have gender info but we can infer it from some tools with degree of accuracy. If we consider the inaccuracies as noise, ...