ProfileXAI: User-Adaptive Explainable AI
Gilber A. Corrales, Carlos Andrés Ferro Sánchez, Reinel Tabares-Soto, Jesús Alfonso López Sotelo, Gonzalo A. Ruz, Johan Sebastian Piña Durán
TL;DR
The paper addresses the challenge of explainability for diverse audiences by proposing ProfileXAI, a domain- and model-agnostic framework that couples post-hoc explainers (SHAP, LIME, Anchor) with retrieval-augmented generation over a multimodal knowledge base. It dynamically selects an explainer per instance using quantitative criteria and tailors narratives via profile-specific prompts in a chat-enabled interface. On Heart Disease and Differentiated Thyroid Cancer Recurrence datasets, the approach achieves stable token usage and positive satisfaction across three user profiles, with LIME offering the best fidelity–robustness balance ($Infidelity$ in $[0.08,0.30]$, $L<0.7$), Anchor delivering sparsity and low token load, and SHAP delivering the highest perceived quality ($\bar{x}\approx 4.0$–$4.1$) at higher complexity. Profile-conditioned explanations demonstrate practical potential for interpretable AI, while future work will extend multimodal support, add more explanation strategies, and pursue human-subject validation.
Abstract
ProfileXAI is a model- and domain-agnostic framework that couples post-hoc explainers (SHAP, LIME, Anchor) with retrieval - augmented LLMs to produce explanations for different types of users. The system indexes a multimodal knowledge base, selects an explainer per instance via quantitative criteria, and generates grounded narratives with chat-enabled prompting. On Heart Disease and Thyroid Cancer datasets, we evaluate fidelity, robustness, parsimony, token use, and perceived quality. No explainer dominates: LIME achieves the best fidelity-robustness trade-off (Infidelity $\le 0.30$, $L<0.7$ on Heart Disease); Anchor yields the sparsest, low-token rules; SHAP attains the highest satisfaction ($\bar{x}=4.1$). Profile conditioning stabilizes tokens ($σ\le 13\%$) and maintains positive ratings across profiles ($\bar{x}\ge 3.7$, with domain experts at $3.77$), enabling efficient and trustworthy explanations.