Fuzzy Logic Enhances Answer Set Programming for Qualitative Reasoning.

Human decision-making often relies on qualitative concepts like "high" or "low," which are inherently vague and context-dependent, even when rooted in numerical data. Traditional symbolic reasoning systems can struggle with this ambiguity, often requiring rigid thresholds that don't reflect human intuition. This research explores a new method to integrate such qualitative reasoning with computational logic. The paper presents a novel extension to Answer Set Programming (ASP) that incorporates fuzzy membership functions. This approach allows for a more nuanced interpretation of linguistic labels, moving beyond binary true/false statements to degrees of truth. The underlying language, detailed in a separate work, provides a principled way to handle vagueness and uncertainty. Through a case study, the framework demonstrates its ability to combine numerically grounded inputs, such as those from machine learning models, with symbolic reasoning over qualitative labels. Key features include learning-based membership functions and semantically enriched predicates, enabling the integration of expert knowledge, contextual factors, and subjective interpretations within a unified declarative programming environment. This offers a robust way to bridge the gap between precise numerical data and imprecise human concepts.