| Nome do aluno | Roberto de Cerqueira Figueiredo |
|---|---|
| Título do trabalho | Intention Model to Provide Pragmatic Interoperability in Collaborative System of Systems |
| Resumo do trabalho |
Context/Problem: Interoperability serves as one of the key building blocks for realizing collaboration in Systems of Systems (SoS). Among its levels of interoperability, pragmatic interoperability aims to ensure that systems can understand one another’s intention, thereby achieving the desired effects. Current solutions do not adequately handle intention interpretation between systems, particularly when dealing with softgoals that underlie decision processes. This gap becomes highly significant in situations like vehicle–driver interaction, where the vehicle system cannot recognize the driver’s system intention and related softgoals, resulting in limited collaborative behavior and adaptation. Aims: This study seeks to: develop an architecture that enables pragmatic interoperability among systems through the interpretation of intentions; incorporate models of intention, namely the BDI (Belief–Desire–Intention) model and the i* Strategic Rationale (SR) model, understood through a Large Language Model (LLM) for reasoning; and validate the proposed architecture in scenarios to determine its ability to achieve pragmatic interoperability through interpretation of intention and softgoals. Method: The research agenda includes: • Architectural Development: Develop an architecture that combines BDI intention models for agent decision-making, i* SR models to represent softgoals and strategic reasoning, and message-exchange middleware between systems. • LLM Integration: Use a Large Language Model to interpret i* SR models and extract contextual and intentional information, enabling reasoning about system intentions. • Scenario-Based Validation: Apply the architecture to simulated scenarios in which messages are communicated between systems, with softgoals represented using i* SR for each scenario. • Evaluation: Measure pragmatic interoperability by analyzing the systems’ ability to share intention understanding, ensuring that message use produces results aligned with expected effects. Expected Results: Successful implementation of the architecture integrating BDI models, i* SR models, middleware, and LLM-based reasoning; validated architecture through simulation demonstrating effective message exchange between systems, with the ability to interpret intentions and softgoals at runtime; contribution to pragmatic interoperability research through providing an intention interpretation mechanism to address known gaps in the existing literature. Future work will explore the explainability quality of the models and how it contributes to achieving pragmatic interoperability. |
| Orientador | Daniela Barreiro Claro |
| Co-orientador | Rita Suzana Pitangueira Maciel |
| Membro externo 1 (com afiliação) | Valdemar Vicente Graciano Neto - Universidade Federal de Goiás (UFG) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Membro interno 1 (com afiliação) | Jose Maria Nazar David - Universidade Federal de Juiz de Fora |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Suplente do membro externo (com afiliação) | julio cesar sampaio do prado leite - Universidade Federal da Bahia (UFBA) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Suplente do membro interno (com afiliação) | REGINA MARIA MACIEL BRAGA VILLELA - Universidade Federal de Juiz de Fora |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Data do exame | 15 Dec, 2025 |
| Horário do exame | 1:00 PM |
Data da Defesa:
15/12/2025 - 13:00
Tipo de Defesa:
Qualificação de Doutorado