UCOT: semiautomatic generation of conceptual models from use case descriptions
SE '08 Proceedings of the IASTED International Conference on Software Engineering
Hi-index | 0.00 |
This paper contributes both to research and practice on process mining. Previous research on process mining has focused on mining patterns from event log files to generate process models. The process mining approach adopted in this paper is focused on producing patterns about process models, not the models themselves. The approach is demonstrated by ProcMiner - an explorative research prototype for management, consolidating, publishing, retrieving, and analyzing process models. Content-based document clustering is applied to process models represented as XML database in order to find topical groups from models. In practice, organizations face numerous challenges in managing their process models. The models may be heterogeneous or ambiguous. The modeling software may change over time or due to differences in departmental purchases. ProcMiner was used in quality system development initiative at the University of Jyväskylä. The findings support previous model engineering research, showing that multiple actions are needed to ensure consistency of process models, and to make them efficiently manageable.