Assisting Process Modeling by Identifying Business Process Elements in Natural Language Texts

Abstract

Process modeling plays a significant role in the business process lifecycle, as it must stress the quality of process models for supporting all the next steps. However, this phase is time consuming and expensive, a consequence of the huge amount of unstructured input information. In a previous research, we presented an approach for identifying business process elements in natural language texts which facilitate the modeler’s work. Such approach relies on a set of mapping rules associated with natural language processing techniques. The identification itself was already validated, but how to apply this information to minimize the modelers’ effort remains unclear. Highlighting the identified rules in the text can enhance its comprehensibility. This paper explores the applicability of such mapping rules on supporting the modeler by marked up texts. The validation shows promising results, as the time spent and effort perceived by the modeler were both minimized.