Decision Helper Weight Elicitation API

API for calculating evaluation criteria weights using multi-criteria decision-making methods, specifically the Analytic Hierarchy Process (AHP) and Best-Worst Method (BWM). Developed as part of the GEDAI-UFAL project: "Ferramentas Baseadas em Agentes Inteligentes de Apoio à Decisão para Seleção de Projetos de Negócios Inovadores em Pré-Incubadoras e Incubadoras."

---

Project

This backend service provides an interface for decision support processes, allowing agencies and evaluators to:

• Manage and evaluate innovation projects.
• Calculate criteria weights using AHP and BWM methods.
• Integrate these weights into multi-agent systems for innovation project evaluation.

---

⚙️ Technologies Used

• Python 3.11
• FastAPI
• pyDecision
• Docker

---

📄 API Endpoints

The API is organized into two main sections: Projects for managing decision problems and Weights for calculating criteria weights using specific methods.

1. Projects Endpoints

These endpoints manage the core project entities, which consist of alternatives and criteria.

• GET /projects/

  • Description: Retrieve all evaluation projects.

• POST /projects/

  • Description: Create a new evaluation project.

• GET /projects/{evaluation_id}

  • Description: Retrieve a specific evaluation project by its unique ID.

• DELETE /projects/{evaluation_id}

  • Description: Delete an evaluation project.

• PUT /projects/{evaluation_id}

  • Description: Update an entire evaluation project.

• POST /projects/{evaluation_id}/alternatives/

  • Description: Add a new alternative to an existing project.

• POST /projects/{evaluation_id}/criteria/

  • Description: Add a new criterion to an existing project.

2. Weights Endpoints

These endpoints are specifically for calculating and retrieving criteria weights using the AHP and BWM methods.

• POST /weights/ahp

  • Description: Retrieve criterion weights from the AHP method.
  • Request Body Example:

    {
      "criteria": ["Cost", "Impact", "Risk"],
      "matrix": [
        [1, 2, 3],
        [0.5, 1, 4],
        [0.33, 0.25, 1]
      ]
    }

• POST /weights/bwm

  • Description: Retrieve criterion weights from the BWM method.
  • Request Body Example:

    {
      "criteria": ["Cost", "Inovation", "Tecnology", "Management", "Market"],
      "most_important": [2, 1, 4, 3, 8],
      "least_important": [4, 8, 2, 3, 1]
    }

---

📚 References

• Saaty, T. L. (1990). The Analytic Hierarchy Process. McGraw-Hill.
• Rezaei, J. (2015). Best-Worst Method: A new approach to multi-criteria decision-making. Omega, Volume 53, Pages 49–57. https://doi.org/10.1016/j.omega.2014.11.009
• pyDecision Documentation: https://pypi.org/project/pyDecision/
• FastAPI Documentation: https://fastapi.tiangolo.com/