Privacy and Transparency in Graph Machine Learning: A Unified Perspective

TL;DR

This paper offers a unified perspective on the interplay between privacy and transparency in Graph Machine Learning, highlighting challenges and future research directions for trustworthy AI in sensitive applications.

Contribution

It introduces a formal framework to study privacy-transparency tradeoffs in GraphML, addressing a gap in integrated understanding of these issues.

Findings

Identifies key challenges in privacy and transparency for GraphML

Proposes research directions for formal analysis of tradeoffs

Highlights importance for trustworthy AI in sensitive domains

Abstract

Graph Machine Learning (GraphML), whereby classical machine learning is generalized to irregular graph domains, has enjoyed a recent renaissance, leading to a dizzying array of models and their applications in several domains. With its growing applicability to sensitive domains and regulations by governmental agencies for trustworthy AI systems, researchers have started looking into the issues of transparency and privacy of graph learning. However, these topics have been mainly investigated independently. In this position paper, we provide a unified perspective on the interplay of privacy and transparency in GraphML. In particular, we describe the challenges and possible research directions for a formal investigation of privacy-transparency tradeoffs in GraphML.