Fuzzy Causal Loop Diagrams: Structure-Informed Diagrams Enabling Qualitative Information for Computational Analysis of Systems

Systems are sets of interacting elements whose collective behavior emerges from their relationships rather than from the behavior of individual components. Systems science employs various mapping tools to understand system dynamics, particularly Causal Loop Diagrams (CLDs) and Fuzzy Cognitive Maps (FCMs). CLDs excel at qualitative system mapping with stakeholders, while FCMs offer computational estimation capabilities through weighted causal relationships. This article proposes the Fuzzy Causal Loop Diagram (FCLD), a method that bridges CLDs and FCMs. The approach estimates edge weights using a modified random walk algorithm. The weight is interpreted as the importance of causal relationships. This innovation enables computational analysis of qualitative stakeholder knowledge, expanding applications to knowledge-based system estimation. The method was tested on two CLDs converted to FCLDs, with validation through weight stability assessment and leverage point analysis. Results show stable weight attribution and leverage point estimates that largely align with CLDs, while offering more informed predictions, particularly for larger systems. This approach opens new possibilities for designing knowledge-based computational systems based on qualitative stakeholder input.