Reference modeling provides significant advantages for both research and practical application. In doing so, considerable challenges arise particularly in the integration of large quantities of process models.
One approach to address these challenges is the inductive reference modeling, which is based on a generalization of business models. Inductive reference modeling concentrates on the similarities between the business models and abstracts them by means of specific features („bottom-up“ approach).
Inductive reference modeling can be divided into six successive steps:
- Preparation: Before a reference model can be developed, several decisions and assumptions regarding the context of the modeling have to be made. Key issues include, for example, the identification of stakeholders‘ demands and the underlying conventions of a modeling.
- Collection: In this step, the individual models required for inductive reference modeling are collected and sorted.
- Preprocessing: The collected individual models must then be adjusted and harmonized in order to enable a subsequent abstraction.
- Acquisition: The prepared individual models are used to derive a corresponding reference model. This is done by arbitrary (?) similarity measures and construction approaches.
- Post Processing: After the previous automatic processing, any errors occurring in the reference model (unconnected model parts, incorrect or unnamed nodes etc.) are now manually corrected.
- Evaluation and Enhancing: In a continuous process, the reference model is regularly evaluated and new circumstances are included.
How the RefMod-Miner works:
The RefMod-Miner provides all the basic functions required to perform inductive reference modeling. For the preparation of the respective model and the subdivision into individual models as well as for the concrete development of a reference model, integrated functionalities can be found.
Furthermore, a detailed follow-up of the developed reference model is made possible.