Title: Causality in Decision Making: The R.Graph Perspective
Abstract.
Causal models play a vital role in multiple criteria decision making (MCDM), particularly in environments characterized by complexity, risk, and uncertainty. These models enable decision-makers to identify, represent, and analyze the dynamic interactions between various criteria, objectives, and alternatives. This capability is especially crucial in industrial domains, where interdependencies between system components can significantly influence both short-term operations and long-term strategic planning. A diverse set of tools has been developed to support causal modeling in decision science, each offering unique strengths and facing certain limitations. In this presentation, we begin by briefly reviewing prominent causal modeling techniques used in MCDM. These include the Decision-Making Trial and Evaluation Laboratory (DEMATEL), Fuzzy Cognitive Maps (FCMs), Analytic Network Process (ANP), and Bayesian Networks. Following this overview, we turn our attention to a novel approach known as R.Graph, a recently developed model designed to enhance decision-making processes under uncertainty. The R.Graph methodology offers a flexible and powerful framework for capturing both direct and indirect interactions among decision elements. It supports the integration of expert knowledge, data-driven insights, and subjective judgments, making it highly adaptable to a wide range of decision environments. We delve into the theoretical foundations of the R.Graph model, illustrating how it can be applied to both short-term operational decisions and long-term strategic planning which positions R.Graph as a promising tool for resilience-oriented and adaptive decision making. Furthermore, we demonstrate how R.Graph can function as a multi-attribute decision-making tool by capturing complex interdependencies not only between criteria but also among alternatives—or simultaneously between both. Finally, we explore how the R.Graph approach could support multi-objective decision making.
