Documento de conferencia
Acceso Abierto

Specification Cases: a lightweight approach based on Natural Language


Use Cases are one of the most used artifacts in software requirements specifications. Although there are a lot of templates that suggests how to describe Use Cases, as well as many quality inspection techniques, there are no many techniques to deal with the complexity and the effort to produce good quality Use Cases. On top of that, Use Cases are biased towards user interaction, leaving out important domain information. Thus, it is necessary to provide techniques to incrementally describe specifications that goes beyond user interaction but which can be derived from simpler artifacts in order to make the process organized and effective. This paper proposes an approach to begin with very simple sentences (kernel sentences), following with the LEL glossary and Scenarios, in order to describe Specification Cases. The approach relies on already defined kernel sentences and suggests three procedures to reorganize the knowledge captured in kernel sentences to describe the LEL, then Scenarios and finally Use Cases. This paper also reports a preliminary evaluation that supports the applicability and usability of the approach.

Palabras clave
Use Cases
Kernel Sentences

Esta obra se publica con la licencia Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 Internacional
Imagen en miniatura