Forschungsthemen
[DA] Modeling ERP systems with Roles and Collaborations
Nowadays, business enterprise application development is very complex and expensive. Many different organization structures (actor roles), development process methodologies (iterative, agile), and technologies (modeling concepts, CASE tools) try to optimize development processes and maximize product quality in order to satisfy stakeholder needs and requests.
Basically, everything can be traced back to the principle “Divide and Conquer“, also known as Separation of Concerns. While many approaches, such as AOP, achieve the separation of non-functional requirements from functional requirements at design, implementation, and deployment time, there seems to be no meaningful concept or technology for the separation of functional requirements in domain models of enterprise applications.
Business objects that are involved in many functional requirements or collaborations tend to be complicated and hard to understand, adapt, maintain, and reuse. Furthermore, different collaborations change the involved business objects in parallel, causing many update problems. Therefore, a separation of functional requirements in business objects would lead to several benefits:
1. Data, Behavior, and Interaction Modeling of Business Objects:
The business class behavior in each collaboration can be independently developed, tested, and deployed.
2.Runtime Concurrency Behavior
Different collaborations can change the state of a business object at the same time without causing update problems.
3. Development Life Cycle
The delivery or update of a package that adds responsibilities to business objects that already exist would not affect the existing packages.
The goals of this thesis are:
1. Give an overview of up-to-date process methodologies and technologies used for business enterprise application development processes.
2. Introduce a motivating existing case study.
3. Analyze approaches that separate class behavior (role modeling, aspect-oriented programming).
4. Select or develop a new Separation of Concern approach suitable for the problem and apply it to improve the motivating case study.
5. Compare the original with the improved case study.
Betreuer: Jendrik Johannes