Requirements Engineering at a Glance: Comparing GORE and UML methods in the design of automated systems
Javier Martinez Silva1; José Reinaldo Silva1; Arianna Zoila Olivera Salmon2; Pedro Manuel González del Foyo3
1 Universidade de São Paulo; 2 Universidade de Pernambuco; 3 Universidade Federal de Pernambuco
Baixar PDF doi:10.20906/CPS/CBA2018-0644
Resumo
The early phase of Requirements Engineering is crucial to design intelligent automated systems, especially if an analytic formal solutions are not achieved. In such cases Requirement analysis is not feasible and the design process will be more suitable do re-work. Planning and scheduling problems can be approached with Artificial Intelligence (AI) and constitute an important area of machine intelligence - together with machine learning. Therefore, it will be an important result to propose good design practices to solve planning problems, also giving special attention to the early requirements phase. Specifically we would look for a method that lead to formal methods of requirements analysis using Petri Nets (which is also used to direct solutions that do not use AI). In this work we compare two direct approaches to requirements engineering: A semi-formal modeling using Unified Modeling Language (UML) that transform the diagrams to Petri Nets to perform analysis, and another approach that uses Goal-Oriented Requirement Engineering (GORE) method, represented by KAOS diagrams that could also be converted to Petri Nets. The possibility to use a unified Petri Net environment and introduce some extensions such as hierarchy is also envisaged.
Palavras-chave: Petri Net; Requirements Engineering; Intelligent Automated Planning; GORE; UML