A clustering approach based on a quadripartitioned neutrosophic minimal spanning tree and its application using probabilistic membership values

The paper proposes an algorithm with an aim to define a preferencing order on clusters obtained by partitioning quadripartitioned neutrosophic graphs (QNGs). For this purpose, a network system is represented as a QNG and a new probabilistic approach has been proposed in deducing the membership values corresponding to the degrees of the vertices. The technique to determine the minimal spanning tree (MST) using cost and benefit score functions is introduced. The MST, thus obtained, is used to partition the graph into multiple clusters (as viable). Finally a preferencing order is defined on the clusters so isolated. An application in the form of designing a plan for running customized advertisements on social media platforms depending on an individual’s online engagement data is shown using the proposed technique.