Towards Building a Next-Generation Data Analytics Toolbox: Application of the Axiomatic Theories Fusion Methodology

Abstract

Reliable development of next-generation data analytics toolboxes (N-GDATs) requires robust underpinning theories, which cannot necessarily be inductively generated. Based on the axiomatic theories fusion (ATF) methodology we deductively developed a comprehensive theory supporting the development of a N-GDAT for white goods design based on middle-of-life data. Accordingly, theories about designer’s needs, advanced technologies, data analytics, creative problem-solving, decision-making, and interoperability were fused following the ATF steps: (i) selection of component theories, (ii) axiomatic discretization of foundational theories, (iii) establishing relationships among axioms and postulates, (iv) transcription of system of axiomatic propositions into a textual format, and (v) validation of explanatory theory. The obtained new theory provides a robust basis for the targeted knowledge platform. It provides (i) decision- making, (ii) algorithmic concepts, (iii) learning, (iv) data management, (v) interfacing, (vi) reasoning, (vii) data types and characteristics, (viii) design issues, (ix) analytics techniques and methods, and (x) outputs requirements to develop N- GDATs.