Abstract
Modeling of functional requirements of an information system is an important task, and there are
numerous methods for doing so. We compared two alternative modeling methods for defining
functional requirements in object-oriented information system development process: one is OO-DFD transactions - part of FOOM methodology; the other is Use Cases (UC) - part of the UML. The two methods consist of diagrams followed by descriptions that add explanations to the diagrams.
In a controlled experiment subjects performed two tasks: a) comprehension of models; b) creation of models. In the first task, each subject was given a set of diagrams and descriptions of a certain system that has been modeled with one of the two modeling methods, and was asked questions aimed to test how well the model is understood. In the second task, each subject was given a narrative requirements document and was asked to create appropriate models using one of the modeling methods.
One result of the experiment was that the quality of models created with the UC method is
significantly better than those created with the alternative method, though a closer examination of the results showed that the quality for the UC method is better only in one aspect of the method. On the other hand, there was no significant difference in comprehension of the two models. We concluded that while the OO-DFD transaction diagrams are more informative, their descriptions are overly detailed and structure. On the other hand, the UC diagrams are not sufficiently informative. Based on that, we propose an improved method to model functional requirements – Enhanced Use Cases –which combines features from the two explored methods.
numerous methods for doing so. We compared two alternative modeling methods for defining
functional requirements in object-oriented information system development process: one is OO-DFD transactions - part of FOOM methodology; the other is Use Cases (UC) - part of the UML. The two methods consist of diagrams followed by descriptions that add explanations to the diagrams.
In a controlled experiment subjects performed two tasks: a) comprehension of models; b) creation of models. In the first task, each subject was given a set of diagrams and descriptions of a certain system that has been modeled with one of the two modeling methods, and was asked questions aimed to test how well the model is understood. In the second task, each subject was given a narrative requirements document and was asked to create appropriate models using one of the modeling methods.
One result of the experiment was that the quality of models created with the UC method is
significantly better than those created with the alternative method, though a closer examination of the results showed that the quality for the UC method is better only in one aspect of the method. On the other hand, there was no significant difference in comprehension of the two models. We concluded that while the OO-DFD transaction diagrams are more informative, their descriptions are overly detailed and structure. On the other hand, the UC diagrams are not sufficiently informative. Based on that, we propose an improved method to model functional requirements – Enhanced Use Cases –which combines features from the two explored methods.
| Original language | American English |
|---|---|
| Title of host publication | Proceedings of the 6th Mediterranean Conference on Information Systems (MCIS) |
| Pages | 1-13 |
| State | Published - 2011 |
| Event | Proceedings of the 6th Mediterranean Conference on Information Systems (MCIS) - Limosol, Cyprus Duration: 3 Sep 2011 → 5 Sep 2011 |
Conference
| Conference | Proceedings of the 6th Mediterranean Conference on Information Systems (MCIS) |
|---|---|
| Country/Territory | Cyprus |
| City | Limosol |
| Period | 3/09/11 → 5/09/11 |