A concept of direct knowledge acquisition for multi-agent design systems
Year: 2013
Editor: Udo Lindemann, Srinivasan V, Yong Se Kim, Sang Won Lee, John Clarkson, Gaetano Cascini
Author: Kratzer, Martin; Crostack, Alexander; Kadadihi, Reem; Rauscher, Michael; Binz, Hansgeorg; Goehner, Peter
Series: ICED
Institution: 1: Institute for Engineering Design and Industrial Design (IKTD), University of Stuttgart, Germany; 2: Institute of Industrial Automation and Software Engineering (IAS), University of Stuttgart, Germany
Page(s): 219-228
ISBN: 978-1-904670-49-0
ISSN: 2220-4334
Abstract
In this contribution, the paradigm of direct knowledge acquisition as a possibility to enable experts to integrate user-specific knowledge is applied to multi-agent design systems (MADS). A pure adoption of existing approaches is not sufficient due to the fact that MADS have a distributed knowledge base as opposed to centralised knowledge bases of other knowledge-based systems (e.g. case-based systems). The result is a concept which brings the intentions of design engineers by modifying the knowledge base (e.g. adding a rule) in accordance with the inherent processes in the knowledge integration component of MADS. Thus, an action of a design engineer is followed by certain operations which have been inherently carried out in order to realise the modification of the knowledge base and to keep the knowledge base consistent. In order to do so, an overview about the direct knowledge acquisition is outlined. Moreover, the structure of the distributed knowledge base is presented wherein the ProKon-knowledge forms are used. Finally, the functionality of the direct knowledge acquisition is presented in detail using the example of a key connection.
Keywords: Multi-agent design systems, direct knowledge acquisition, knowledge management