Ermittlung von Konfigurationsprofilen aus Auftragsstücklisten mittels Data-Mining
Editor: Dieter Krause, Kristin Paetzold, Sandro Wartzack
Author: Kevin, Schmitz; Georg, Jacobs; Johannes, van der Beek
Institution: RWTH Aachen
Section: Data Mining
In order to be successful on the market in the long term, it is necessary to offer market-oriented product variants. The first step is to find out which vari-ants of a product the market demands in order to ensure an appropriate ex-ternal diversity. The present paper shows that necessary knowledge in com-panies is implicitly present in the bills of material (BOMs) of the past and can be exploited for the product development by suitable analysis methods. For this purpose, BOM analyses are carried out on the basis of data mining meth-ods. By linking physical product components and functions to a product archi-tecture, complete configuration profiles can be created for all products of the past. They can be used to predict the variant diversity in early development phases in order to reduce it systematically.