Reduction of Cognitive Overload in Online Reviews Using Data Visualisation

Abstract

With the internet taking over many aspects of our lives including the way commercial practices are handled, many business owners are taking what has been posted about them in the forms of online reviews very seriously. While most e-business data visualisation tools focus on website analytics and customer behaviors to determine what customers want and what product or service needs improvement, many dismiss the importance of online reviews, especially the ability to determine if a review should be considered valuable or not. One of the problems of efficiently understanding online reviews is the reader not having enough cognitive strength and working memory to decide if the reviews should be taken seriously or not, and at the same time, understand what the review is about. This paper proposes a novel model to automatically mine online reviews from certain websites, analyses them using decision-tree machine learning and n-grams, and then display a visualisations to highlight how true a review is to be considered. To achieve this, several stages take place in the visualisation system's framework, including retrieving and processing the data, and creating the visualisation. In this study, we focus on reducing cognitive overload by performing some pilot usability studies so that business owners can make better informed decisions.