Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Learning from the past: uncovering design process models using an enriched process mining

Lan, Lijun, Liu, Ying ORCID: and Lu, Wen Feng 2018. Learning from the past: uncovering design process models using an enriched process mining. Journal of Mechanical Design 140 (4) , 041403. 10.1115/1.4039200

[thumbnail of md_140_04_041403.pdf]
PDF - Published Version
Available under License Creative Commons Attribution.

Download (5MB) | Preview


Design documents and design project footprints accumulated by corporate IT systems have increasingly become valuable sources of evidence for design information and knowledge management. Identification and extraction of such embedded information and knowledge into a clear and usable format will greatly accelerate continuous learning from past design efforts for competitive product innovation and efficient design process management in future design projects. Different from existing systems, this paper proposes a methodology of learning and extracting useful knowledge using past design project documents from design process perspective based on process mining techniques. A new process mining approach that is able to directly handle textual data is proposed at the first stage of the proposed methodology. The outcome is a hierarchical process model that reveals the actual design process hidden behind a large amount of design documents and enables the connection of various design information from different perspectives. At the second stage, the discovered process model is further refined to learn multi-faceted knowledge patterns by applying a number of statistical analysis methods. The outcomes range from task dependency study from workflow analysis, identification of irregular task execution from performance analysis, cooperation pattern discovery from social net analysis, to evaluation of personal contribution based on role analysis. Relying on the knowledge patterns extracted, lessons and best practices can be uncovered which offer great support to decision makers in managing any future design initiatives. The proposed methodology was tested using an email dataset from a university-hosted multi-year multidisciplinary design project.

Item Type: Article
Date Type: Publication
Status: Published
Schools: Engineering
Publisher: ASME
ISSN: 1050-0472
Date of First Compliant Deposit: 5 February 2018
Date of Acceptance: 16 October 2017
Last Modified: 05 May 2023 20:30

Citation Data

Cited 6 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics