Mining staff assignment rules from event-based data

  • Authors:

  • Linh Thao Ly;Stefanie Rinderle;Peter Dadam;Manfred Reichert

  • Affiliations:

  • Dept. DBIS, University of Ulm, Germany;Dept. DBIS, University of Ulm, Germany;Dept. DBIS, University of Ulm, Germany;IS Group, University of Twente, The Netherlands

  • Venue:
  • BPM'05 Proceedings of the Third international conference on Business Process Management
  • Year:
  • 2005

Quantified Score

Hi-index 0.01

Visualization

Abstract

Process mining offers methods and techniques for capturing process behaviour from log data of past process executions. Although many promising approaches on mining the control flow have been published, no attempt has been made to mine the staff assignment situation of business processes. In this paper, we introduce the problem of mining staff assignment rules using history data and organisational information (e.g., an organisational model) as input. We show that this task can be considered an inductive learning problem and adapt a decision tree learning approach to derive staff assignment rules. In contrast to rules acquired by traditional techniques (e.g., questionnaires) the thus derived rules are objective and show the staff assignment situation at hand. Therefore, they can help to better understand the process. Moreover, the rules can be used as input for further analysis, e.g., workload balance analysis or delta analysis. This paper presents the current state of our work and points out some challenges for future research.