🌍 Data Analytics for International Political Risk Assessment

Repository for a master’s thesis on political risk and business impacts. It holds data, analysis, and the systematic literature review outputs.

---

🧭 Overview

• Monthly time series of conflict events by city.
• Exploratory and risk analysis (Probability × Impact) in the main notebook.
• Systematic Literature Review (SLR) with search, extraction, and protocol files.

---

🗂️ Folder structure

data/
   gpt.xlsx                  # Monthly conflict series
results/
   dissertation.ipynb        # Notebook with cleaning, analysis, and charts (Colab/Jupyter)
   fluxo_visualizacao.png    # Visualization flow
   graph_*                   # Exported charts (frequency, causes, risk, etc.)
   methodology.png           # Methodology diagram
   slr/
      data/
         scopus.bib
         springer.bib
         springer.csv
         wos.bib
      results/
         # Outputs from Parsif.al and protocol
README.md

---

🚀 How to run the analysis

1. Local environment or Colab

• Open results/dissertation.ipynb in Jupyter/Colab.
• Minimum dependencies:

pip install pandas seaborn matplotlib numpy scikit-learn plotly pycountry openpyxl

• Keep the file data/gpt.xlsx inside the data/ folder (already in the repo).

2. Notebook flow

• Loads and cleans the series (dates, name normalization, “X” field binarization).
• Computes Probability × Impact, top frequencies by capital, and aggregated causes (barplots and heatmaps).
• Exports charts and derived files in the results/ folder.

---

📄 Systematic Literature Review (SLR)

• Sources: Scopus, Springer, Web of Science (files .bib, .csv).
• Folder results/slr/data keeps inputs; results/slr/results stores Parsif.al outputs and the protocol.
• Document results/slr/results/rsl.pdf describes criteria, search, and extraction.

---

📚 Citation

If this work helps, cite:

Sousa, C. A. (2025). Data Analytics for International Political Risk Assessment. Institute of Computer Science. University of São Paulo.