Synthetic Attribute Data for Evaluating Consumer-side Fairness

TL;DR

This paper introduces the FLAG algorithm to generate synthetic demographic attributes for recommender system datasets, enabling fairness evaluation without relying on sensitive real demographic data.

Contribution

The paper presents a novel algorithm for creating synthetic demographic attributes, facilitating fairness analysis in recommender systems without compromising privacy.

Findings

FLAG effectively generates realistic demographic attributes

Enables fairness evaluation without sensitive data

Supports experimentation with different demographic distributions

Abstract

When evaluating recommender systems for their fairness, it may be necessary to make use of demographic attributes, which are personally sensitive and usually excluded from publicly-available data sets. In addition, these attributes are fixed and therefore it is not possible to experiment with different distributions using the same data. In this paper, we describe the Frequency-Linked Attribute Generation (FLAG) algorithm, and show its applicability for assigning synthetic demographic attributes to recommendation data sets.