Facilitating the Interoperability and Reuse of Extensions of Fuzzy Cognitive Maps

Fuzzy Cognitive Maps (FCMs) are widely used to support participants in externalizing their mental models regarding a system. However, the graph-based formalism of FCMs cannot represent nonlinear relationships or time delays/lags and have a limited representation of uncertainty, hindering their ability to model complex systems. Indeed, a relationship’s strength may depend on the originating node (e.g., weight-based discrimination intensifies for obese individuals), while other relationships have a gradual effect (e.g., the effect of investments onto educational attainments may take years) or are highly variable (e.g., stress may lead to eating more or less). Numerous extensions have consequently been proposed (e.g., Fuzzy Grey Cognitive Maps). Supporting interoperability among extensions would facilitate the reuse of models. We developed 26 algorithms that allow modelers to transform models from 10 FCM extensions. Our open-source framework finds the most efficient sequence of algorithms when transforming one framework to another with no direct relation.