Two of the main needs when developing an information system for an organization are that system analysts know and understand the application domain and that the system properly supports the business processes of the organization. Consequently, elicitation of system requirements from business process models has been acknowledged as a suitable activity to deal with that needs. In addition, system requirements must be linked to subsequent development stages.
However, system analysts can face many challenges when performing these activities. They may have problems communicating with customer stakeholders and may need to analyse and operationalize the purpose of the information system. Furthermore, system analysts must bridge the gap between business and system domains for specification of system requirements, specify different types of system requirements and guarantee that their specification is precise, consistent and homogeneous. In relation to object-oriented conceptual modelling-based information system development, system analysts must also avoid potential problems that may arise when a conceptual schema is created from system requirements as part of their link with subsequent development stages. For example, a conceptual schema can be incomplete or inconsistent if it is not properly managed.
As a solution, this thesis presents a methodological approach for business process-based requirements specification and object-oriented conceptual modelling of information systems. The approach consists of four stages: organizational modelling, purpose analysis, specification of system requirements and derivation of object-oriented diagrams.
By following the design research methodology for performing research in information systems, the methodological approach has been designed on the basis of many existing ideas and principles in academia and industry. It also provides new mechanisms and guidance to address the challenges presented above. The methodological approach mainly aims to help system analysts to elicit system requirements from business process models, adequately specify system requirements and derive the object-oriented conceptual schema of an information system from its system requirements. It does so by taking advantage of existing solutions and by modifying them to better tackle the associated challenges.
The methodological approach has been evaluated in laboratory and industrial contexts, especially focusing on its usefulness for practitioners. Thanks to evaluation, several lessons have been learned and many of them have driven definition of the methodological approach. Furthermore, the lessons learned can be very useful both in academia and in industry for identification of further research areas and for awareness of situations that may occur in information system development projects, respectively.