The Business Insight Index – Evaluating Customer Insights through Hybrid Models

Abstract: Customer segmentation and target analysis are two essential tasks when identifying a company’s customers. To perform these tasks, this thesis develops and applies hybrid data-mining models, integrating clustering and decision trees. The hybrid models are applied to the life-logging camera company Narrative, in order to gain insights into their customer data. From previous research, we found that these hybrid models lacked means for evaluating the amount of insights proposed to decision makers. For this reason, we created, tested, and validated a new evaluation measure – the Description Tree Index. Through experiments on five separate datasets, we conclude that the measure enables decision makers to evaluate the insights gained through the hybrid model. In each case, the index generates the best results for the expected number of segments. We then integrated the Description Tree Index with existing evaluation models to form a Business Insight Index. This index evaluates customer segmentation and target analysis from both a business and data-mining perspective. By applying the index to the Narrative data, we found four customer segments to present the most insights.