Recommendation functionality for smart data analytics toolbox to support choosing task-relevant data analytics tools

Conference paper (2020)

Authors

F. Abou Eddahab-Burke Cyber-Physical Systems - IDE
I. Horvath Cyber-Physical Systems - IDE
Research Group
System Engineering (TPM) (TU Delft)
Copyright: © 2020 F. Abou Eddahab-Burke, I. Horvath
Data Analytics White Goods Smart Data Analytics Toolbox Task-relevant Recommendation Recommendation Systems Machine Learning Tools Product Enhancement Support for Designers
More Info
expand_more

Published Date

2020

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Technology, Policy and Management

Department

Multi Actor Systems

Research Group

System Engineering

Abstract

Though many enhancements are still possible and needed, data analytics software packages invaded all segments of industrial businesses. Since product designers are not specialized data analysts, an op- portunity of enhancement is to provide advice by smart data analytics toolboxes (SDATBs). For in- stance, SDATBs can provide guidance at selecting commercially available data analytics tools (DATs) for a specific design-related task. The reported work focused on the implementation of a recommendation functionality for selecting DATs for different appli- cations. The paper presents the proposed solution, which (i) interprets the designer’s input, (ii) pro- poses a description of the problem identified by the designer, (iii) reasons with the warehoused DATs and (iv) recommends DATs matching the designer’s task at hand. Besides presenting the needed func- tionality, the rules used for selecting DATs are dis- cussed and the computational algorithms are speci- fied. A computational feasibility testing of the tool recommendation functionality has been done considering the application case of enhancing a wash- ing machine by white goods designers. The testing process showed that the realized functionality works correctly from a computational point of view and that it achieves sufficiently good tool matching. It compensates for the knowledge lack of product de- signers concerning selection of data analytics tools and reduces time and effort for tools selection. The outcomes of this study will be used in a follow up research to develop a SDATB providing even more comprehensive support for product designers.