From Indicators to Predictive Analytics: A Conceptual Modelling Framework

Abstract

Predictive analytics provides organisations with insights about future outcomes. Despite the hype around it, not many organizations are using it. Organisations still rely on the descriptive insights provided by the traditional business intelligence (BI) solutions. The barriers to adopt predictive analytics solutions are that businesses struggle to understand how such analytics could enhance their existing BI capabilities, and also businesses lack a clear understanding of how to systematically design the predictive analytics. This paper presents a conceptual modelling framework to overcome these barriers. The framework consists of two modelling components and a set of analysis that systematically 1) justify the needs for predictive analytics within the organisational context, and 2) identify the predictive analytics design requirements. The framework is illustrated using a real case adopted from the literature.

A Methodology for Knowledge Discovery in Labeled and Heterogeneous Graphs

“…Nasiri et al [26] presented a framework with two modeling components: context modeling and analytics design alternatives modeling. The first is to justify the need for predictive analytics within the organizational context.…”