Aggregating Procedure for Fuzzy Cognitive Maps

Abstract

In the field of Knowledge Engineering and Representation, a typical struggle encompasses transferring the Subject Matter’s expertise into computational descriptions that could be used to create digital-twin representations of a given real-world scenario. Fuzzy Cognitive Maps (FCMs) have recently gained relevant attention among multiple techniques developed with this aim. However, one issue remains, when numerous of these representations need to be combined into a unique aggregated structure, it is essential to weigh factors (such as quality) into the final form to ensure its veracity, making the process not too straightforward. This paper proposes an aggregation procedure to combine FCMs into one that represents best its contributors. The technique was utilized for solving a real-life problem, and several configurations were explored. The results are compiled and reported in this paper.