article · Journal of Future Artificial Intelligence and Technologies
DiFACE2 is a machine learning framework designed to generate diverse, feasible, and actionable counterfactual explanations using causal Directed Acyclic Graphs. Built with background knowledge from an educational intervention programme in post-conflict Nigeria, the framework identifies causal relationships and immutable variables to explain predictions from black-box models. It was evaluated on the Strengthening Education in Northeast Early Grade Reading Assessment dataset, producing up to four distinct counterfactuals per prediction. While numeric variables occasionally require additional constraints to prevent infeasible suggestions, categorical variables show high feasibility and practical actionability. Although predictive accuracy scores are lower than those achieved on simplified benchmark datasets due to real-world noise, DiFACE2 prioritises practical interpretability and outperforms previous methods across all counterfactual quality metrics when tested on the same complex dataset.
Machine learning models often operate as black boxes, making it difficult to understand how individuals can change inputs to achieve better outcomes. By embedding real-world causal rules into the explanation process, this approach ensures recommendations are realistic rather than merely mathematically possible. This helps decision-makers in complex environments, such as educational programmes, identify practical steps to improve results.
This method is applicable to decision-support software for administrators and analysts managing educational interventions or social programmes. It enables users to understand which operational changes could realistically alter assessment outcomes. The technology is applied and tested, having been validated on a real-world dataset from Nigeria, but remains at an academic prototype stage prior to commercial software integration.
AI-generated from the published abstract. Always read the original work before citing.
We introduce DiFACE2 causal DAG, a framework for generating Diverse, Feasible, and Actionable Counterfactual Explanations under causal Directed Acyclic Graph (DAG) constraints. Leveraging background knowledge from an educational intervention program in post-conflict Nigeria, we construct an application-specific causal DAG to identify causal relationships and immutable features during counterfactual generation for black-box ML predictions. Using this DAG, we develop a causally constrained sparse ML model that integrates prior frameworks to generate diverse, feasible, and actionable counterfactuals for the Strengthening Education in Northeast – Early Grade Reading Assessment (SENSE-EGRA) dataset. DiFACE2 produces up to four counterfactuals per prediction, demonstrating high feasibility and actionability for categorical variables. While numeric features occasionally yield infeasible recommendations, this is mitigated by additional non-causal constraints. Although the sparse ML model yields lower accuracy and F1 scores than benchmark frameworks, this stems from the higher complexity and noisiness of the real-world dataset rather than methodological weakness. This trade-off underscores that meaningful, interpretable counterfactuals are preferable to near-perfect accuracy on simplified benchmarks. Therefore, when fairly evaluated on the same complex dataset, DiFACE2 outperforms prior methods across all counterfactual quality metrics.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.62411/faith.3048-3719-283
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
