NEUTROSOPHIC KNOWLEDGE-BASED FRAMEWORKS FOR INTELLIGENT DECISION SUPPORT UNDER DEEP UNCERTAINTY: APPLICATIONS IN SMART AGRICULTURE

Abstract

Neutrosophic statistics and logic are generalized Fuzzy logic models of uncertainty via an explicit account of terms of truth, indeterminacy and falsity, as extensions of classical probability, fuzzy sets and intuitionistic fuzzy sets. Over recent years, many neutrosophic extensions and hybrid models have been suggested in response to complex decision-making problems involving incomplete, imprecise, and conflicting information. This paper provides a summary of neutrosophic statistical and logical methods, focus on high-level structures, including single-valued, hypersoft, topological, cubic and spherical neutrosophic sets. Mathematical foundations, aggregation operators, and multi-criteria decision-making (MCDM) mechanisms within these frameworks are critically analysed and comparatively discussed. Rapidly evolving formulations in applications, such as multi-criteria and hybrid neutrosophic-learning models are overviewed to demonstrate the ability of neutrosophic models to model hierarchical, multi-attribute, and indeterminate information systems. Smart agriculture has been taken as a representative area of application to illustrate real-world modelling situations without limiting the generality of the offered frameworks. This study proposes a unified conceptual framework integrating neutrosophic theory and multi-criteria decision-making for intelligent decision support under deep uncertainty. The review also presents the major implementation and methodological issues such as parameterization based on experts, absence of standard benchmarks, insufficient software chains, and complexity. Future research directions are provided, including scalable neuro-neutrosophic models, explainable artificial intelligence (XAI) integration, and large-scale decision-support systems.