Towards User-Focused Research in Training Data Attribution for Human-Centered Explainable AI

TL;DR

This paper advocates for a user-centered approach in training data attribution for explainable AI, emphasizing practical needs and real workflows over purely mathematical methods, through interviews and scenario studies.

Contribution

It introduces a design thinking framework for TDA, highlighting novel tasks like grouping data by model behavior and identifying undersampled data, grounded in user studies.

Findings

Identified new data-centric explanation tasks for developers.

Engaged with practitioners to align TDA with real workflows.

Proposed a user-centered research direction for TDA.

Abstract

Explainable AI (XAI) aims to make AI systems more transparent, yet many practices emphasise mathematical rigour over practical user needs. We propose an alternative to this model-centric approach by following a design thinking process for the emerging XAI field of training data attribution (TDA), which risks repeating solutionist patterns seen in other subfields. However, because TDA is in its early stages, there is a valuable opportunity to shape its direction through user-centred practices. We engage directly with machine learning developers via a needfinding interview study (N=6) and a scenario-based interactive user study (N=31) to ground explanations in real workflows. Our exploration of the TDA design space reveals novel tasks for data-centric explanations useful to developers, such as grouping training samples behind specific model behaviours or identifying undersampled data. We invite the TDA, XAI, and HCI communities to engage with these tasks to strengthen their research's practical relevance and human impact.