Em busca de um meta-modelo para a unificação de diferentes abordagens de representação de agentes de software e para geração de código em plataformas de desenvolvimento de sistemas multiagentes

• Ano de defesa: 2009
• Autor(a) principal: Pires, Ivan Luiz Pedroso
• Orientador(a): Ribeiro, Marcelo Blois
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul; Porto Alegre
• Palavras-chave em Português: INFORMÁTICA; SISTEMAS MULTIAGENTES; AGENTE INTELIGENTE (SOFTWARE)
• Link de acesso: http://hdl.handle.net/10923/1472

Multi-agent systems have shown great growth in the area of software development as a promising paradigm to deal with the complexity of current scenarios of information technology. Many approaches arise in order to consolidate ways and means to develop an MAS, which can be classified as methodologies, modeling languages and implementation platforms. In addition, some studies are offered as an attempt to unify the concepts involved and the notations that represent these concepts, with an enormous range of different symbologies, and many concepts in common between these approaches. This paper presents comparisons between these concepts and notations visually in order to find a way to map these approaches, showing their differences and similarities. As a first step towards unification, this work extends the Meta-Model Representation of Internal Agents to allow complete mapping of the Tropos methodology. Thus, we showed how a meta-model can provide interoperability between different approaches severing the creation of a SMA of the requirements of specific methodologies or modeling languages. It's also generated a skeleton source code from this meta-model for SemantiCore framework through a prototype, it is possible to extend it to support the generation of source code for any platform implementation. This process of mapping and code generation is demonstrated by applying an example of the literature.