Transparent Execution of Data Transformations in Data-Aware Service Choreographies

Abstract

Due to recent advances in data science, IoT, and Big Data, the importance of data is steadily increasing in the domain of business process management. Service choreographies provide means to model complex conversations between collaborating parties from a global viewpoint. However, the involved parties often rely on their own data formats. To still enable the interaction between them within choreographies, the underlying business data has to be transformed between the different data formats. The state-of-the-art in modeling such data transformations as additional tasks in choreography models is error-prone, time consuming and pollutes the models with functionality that is not relevant from a business perspective but technically required. As a first step to tackle these issues, we introduced in previous works a data transformation modeling extension for defining data transformations on the level of choreography models independent of their control flow as well as concrete technologies or tools. However, this modeling extension is not executable yet. Therefore, this paper presents an approach and a supporting integration middleware which enable to provide and execute data transformation implementations based on various technologies or tools in a generic and technology-independent manner to realize an end-to-end support for modeling and execution of data transformations in service choreographies.