Abstract
An open challenge of Business Process Management is to obtain a process model (as described) out of a process description document.
Some initial solutions to the problem of process extraction from natural language documents have been proposed in the literature. But, the analysis of the relevant contributions on this topic reveals that the these contributions mainly adopt ad-hoc rule-based approaches and do not scale up to extraction in real world scenarios. Also, the lack of effective validation metrics, a common real world test data sets as well as solid bench-marking procedures make a comparison of the proposed approaches rather difficult. Therefore, process extraction form natural language text can considered an unresolved problem still in an early stage of development. In this paper I want to illustrate limitations of the current approaches in this topic. These open directions are the starting point of my PhD research.