Twin Systems for DeepCBR: A Menagerie of Deep Learning and Case-Based Reasoning Pairings for Explanation and Data Augmentation

TL;DR

This paper explores the integration of Case-Based Reasoning with Deep Learning to enhance explainability and data augmentation, demonstrating twin systems that improve interpretability of opaque models.

Contribution

It introduces twin systems pairing DL models with CBR models for explainability and data augmentation, showcasing novel synergy between these AI paradigms.

Findings

CBR models can effectively explain DL decisions using factual and counterfactual strategies.

Functional abstractions of DL models facilitate explanation generation.

DeepCBR approaches have potential for data augmentation in DL.

Abstract

Recently, it has been proposed that fruitful synergies may exist between Deep Learning (DL) and Case Based Reasoning (CBR); that there are insights to be gained by applying CBR ideas to problems in DL (what could be called DeepCBR). In this paper, we report on a program of research that applies CBR solutions to the problem of Explainable AI (XAI) in the DL. We describe a series of twin-systems pairings of opaque DL models with transparent CBR models that allow the latter to explain the former using factual, counterfactual and semi-factual explanation strategies. This twinning shows that functional abstractions of DL (e.g., feature weights, feature importance and decision boundaries) can be used to drive these explanatory solutions. We also raise the prospect that this research also applies to the problem of Data Augmentation in DL, underscoring the fecundity of these DeepCBR ideas.