Interactive Visualisation of Automotive Warranty Data Using Novel Extensions of Chord Diagrams

Abstract

Descriptive and predictive analytics enable the automotive industry to be pro-active in its management of warrantyrepairs and the substantial unknown costs associated with future claims. Understanding what makes theircustomers most dissatisfied can assist manufacturers to take pro-active steps towards restoring satisfaction andincreasing the likelihood of customers making future purchases. Presented in this short work-in-progress paperare two techniques which enable the interactive visualisation of high-dimensional relationships within categoricaldata. Our Multi-Chord Diagram and Multi-Chord Glyph Diagram extend traditional Chord Diagrams, overcomingthe limitation of only depicting relationships between category pairs. We present these techniques in an application,addressing a real-world problem of visualising data relating to customer satisfaction following variouscombinations of automotive warranty repair. Although presented as a novel application for visualising customersatisfaction, we feel that the techniques described could also be applied to other scenarios involving the visualisationof variable high-dimensional relationships within categorical data.