FairFoody: Bringing in Fairness in Food Delivery

Anjali Gupta,Rahul Yadav,Ashish Nair,Abhijnan Chakraborty,Sayan Ranu,Amitabha Bagchi

DOI: https://doi.org/10.1609/aaai.v36i11.21447

2022-04-25

Abstract:Along with the rapid growth and rise to prominence of food delivery platforms, concerns have also risen about the terms of employment of the gig workers underpinning this growth. Our analysis on data derived from a real-world food delivery platform across three large cities from India show that there is significant inequality in the money delivery agents earn. In this paper, we formulate the problem of fair income distribution among agents while also ensuring timely food delivery. We establish that the problem is not only NP-hard but also inapproximable in polynomial time. We overcome this computational bottleneck through a novel matching algorithm called FairFoody. Extensive experiments over real-world food delivery datasets show FairFoody imparts up to 10 times improvement in equitable income distribution when compared to baseline strategies, while also ensuring minimal impact on customer experience.

Social and Information Networks,Multiagent Systems

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to achieve fair income distribution in food - delivery platforms while ensuring timely food delivery. Specifically, the researchers analyzed the data of real - food - delivery platforms from three large cities in India and found that there is significant inequality in the income among delivery agents. This inequality is not only unrelated to working hours or working periods, but also related to geographical distribution, that is, the order volume in some areas is higher than that in other areas, resulting in higher income for delivery agents working in these areas.

To address this challenge, the paper proposes a new algorithm named FAIRFOODY. This algorithm achieves fair income distribution by optimizing multi - objective problems, that is, minimizing the income gap among delivery agents without sacrificing the customer experience as much as possible. The researchers have proved that this problem is not only NP - hard, but also non - approximable in polynomial time. To solve this computational bottleneck, the FAIRFOODY algorithm adopts a method based on bipartite graph matching and can perform fair order allocation on real - time data streams.

In general, the main contributions of the paper include:
- Conducting in - depth research on real - food - delivery data, revealing that there is also high inequality in the income of delivery agents even after standardizing by active hours.
- Proposing the formulation of multi - objective optimization problems and algorithm design, aiming to achieve fair income distribution in food - delivery tasks.
- Proving through experiments the effectiveness of the FAIRFOODY algorithm on actual food - delivery data, which can significantly improve the fairness of income distribution while maintaining a good customer experience.