Explainable Artificial Intelligence (XAI) has become an important research direction because intelligent systems are increasingly used to support decisions in complex and uncertain environments. However, the interpretability of a decision depends not only on the final prediction but also on the mathematical structure through which the prediction is obtained. This paper develops an algebraic framework for interpretable multi-criteria decision-making based on bounded ordered structures, weighted aggregation, and explicit criterion-level contributions. A finite set of alternatives is evaluated using normalized criteria, and a weighted score is assigned to each alternative. The framework represents criterion values in a bounded lattice and uses an order-preserving aggregation operator to obtain transparent rankings. Mathematical properties concerning monotonicity, boundedness, score decomposition, and stability under perturbations are established. An illustrative numerical example involving the selection of an artificial intelligence model demonstrates how the framework generates rankings and human-readable explanations. The approach provides a mathematically explicit foundation for decision support and can be extended to fuzzy, linguistic, and neutrosophic environments. The proposed framework and numerical example are illustrative; novelty and performance claims require literature verification and empirical validation.

Keywords: Abstract algebra; lattice theory; explainable artificial intelligence; multi-criteria decision-making; interpretable decision systems; weighted aggregation.