Uma regra de sincronização baseada na lógica linear para prevenção de deadlock em workflow nets interorganizacionais

• Ano de defesa: 2017
• Autor(a) principal: Oliveira, Vinícius Ferreira de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Ciência da Computação
• Palavras-chave em Português: Processos de negócio; Business process; Redes de Petri; Petri nets; Lógica Linear; Linear Logic; WorkFlow net Interorganizacional; Interorganizational WorkFlow net; Soundness; Deadlock; Regra de sincronização; Synchronization rule; Computação; Computer Science; Fluxo de trabalho; Workflow; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::METODOLOGIA E TECNICAS DA COMPUTACAO::ENGENHARIA DE SOFTWARE
• Link de acesso: https://repositorio.ufu.br/handle/123456789/21273; http://dx.doi.org/10.14393/ufu.di.2018.1114

This work presents a prevention method for deadlock situations, through a synchronization rule, in interorganizational workflow processes that are locally sound, but not globally sound. Such a method considers the Petri net theory and is based on the analysis of Linear Logic proof trees. The work also presents a method for the qualitative and quantitative analysis of the interorganizational workflow model process, before and after the application of the synchronization rule used to remove deadlock situations. Following the application of the synchronization rule, those models that represent the interorganizational workflow processes that are not sound are transformed into sound models due to the replacement of certain asynchronous communication places by partially synchronous communication mechanisms. As a consequence of the application of the synchronization rule, the process may suffer minor delays, reducing the process performance. The use of Linear Logic has the greatest advantage the possibility of reusing fragments of the proof tree produced before the application of the synchronization rule. Consequently the analysis of the model is only performed in that part of the model responsible for the deadlock situation.