Towards Responsible and Fair Data Science: Resource Allocation for Inclusive and Sustainable Analytics

TL;DR

This paper proposes a novel framework for responsible data science that emphasizes fairness, inclusivity, and sustainability by integrating decolonial perspectives into resource allocation and algorithm design.

Contribution

It introduces new fairness metrics and algorithms that incorporate socio-economic, cultural, and ethical considerations for equitable data science practices.

Findings

Development of fairness metrics respecting cultural diversity

Algorithms optimizing resource allocation for inclusivity

Enhanced transparency and community empowerment in data science

Abstract

This project addresses the challenges of responsible and fair resource allocation in data science (DS), focusing on DS queries evaluation. Current DS practices often overlook the broader socio-economic, environmental, and ethical implications, including data sovereignty, fairness, and inclusivity. By integrating a decolonial perspective, the project aims to establish innovative fairness metrics that respect cultural and contextual diversity, optimise computational and energy efficiency, and ensure equitable participation of underrepresented communities. The research includes developing algorithms to align resource allocation with fairness constraints, incorporating ethical and sustainability considerations, and fostering interdisciplinary collaborations to bridge technical advancements and societal impact gaps. This work aims to reshape into an equitable, transparent, and community-empowering practice challenging the technological power developed by the Big Tech.