Article,
An approach to investigate fairness using Dominance-based Rough Sets Analysis—How fair were the COVID-19 restriction decisions in the UK?
Affiliations
- [1] University of Southern Denmark [NORA names: SDU University of Southern Denmark; University; Denmark; Europe, EU; Nordic; OECD];
- [2] COMSATS University Islamabad [NORA names: Pakistan; Asia, South];
- [3] University of Leeds [NORA names: United Kingdom; Europe, Non-EU; OECD]
Abstract
Fairness is a crucial aspect to consider within decision support systems, to seek to strive for equitable decision outcomes. Therefore, in this work, we introduce an approach to investigate fairness in data-driven decisions. The Dominance-based Rough Sets Approach (DRSA) has been widely used to extract a single set of if-then types of rules from data. Conversely, our approach investigates fairness by extracting multiple separate if-then rule sets for separate groups. The proposed approach facilitates fairness analysis to be performed amongst groups represented by these rule-sets. During the COVID-19 pandemic, several countries have taken the approach of tiered restrictions, which has remained a point of debate due to a lack of transparency. Using our proposed approach, we explore fairness analysis with regards to the UK government’s COVID-19 tiered restrictions allocation system. These insights from the analysis are translated into “if-then” type rules, which can easily be interpreted by policy makers. The differences in the rules extracted from different geographical areas suggest inconsistencies in the allocations of tiers in these areas. We found that the differences delineated an overall north south divide in England, however, this divide was driven mostly by London. Such analysis could provide a more transparent approach to localised public health restrictions, which can help ensure greater conformity to the public safety rules. Our analysis demonstrates the usefulness of our approach, to explore fairness analysis in terms of equal-treatment within data-driven decisions, which could be applied in numerous other domains, for investigating the fairness and explainabilty of decisions.