Enhancing Business Dashboards with Explanatory Analytics & AI : Exploring the Use of AI and Explanatory Analytics to Enhance Business Decision-Making

Abstract

Business dashboards have become increasingly popular for descriptive analytics, significantly aiding the decision-making process. With the growth in data volume and complexity, ex- planatory models for automatic diagnostic analytics have been developed. This thesis aims to enhance these models, thereby extending the functionality of business dashboards by generating textual explanations instead of relying on traditional visualisations or summarised tables. This approach facilitates more informed decision-making. This thesis employs design science research, where an artefact is created based on theory. The research evaluates three theoretical frameworks for diagnostic modelling, ultimately in- corporating only one: the explanation formalism. By utilising generative AI, the developed artefact translates intricate data insights into easily understandable narratives. This approach bridges the gap between data analysis and decision-making, providing clear and comprehensible explanations of data trends and anomalies. The created artefact was tested using a cognitive walkthrough, demonstrating its capability to transform complex data figures into comprehensible text format. Additionally, a comparison was made between the outputs of explanation formalism, an explanatory tree, and explanatory text. This thesis contributes to the fields of information management and business intelligence by presenting a method to integrate existing explanatory models with AI. The result is a prototype that enhances the clarity and understanding of business data.