Yaw Assensoh Opoku

20th Century Geopolitical Network Analysis

Advanced Network Science & Graph Theory Analysis of International Relations

Phase 1

Data Collection

Phase 2

Text Mining & NLP

Phase 3

Network Analysis

20th Century Geopolitical Network Analysis

Project Overview

Goal:

Apply network science and graph theory to analyze 20th century international relationships, identifying geopolitical communities, central powers, and bridge nations using advanced centrality measures.

Business Problem:

Understand the complex web of international alliances and power dynamics that shaped 20th century geopolitics through quantitative network analysis and community detection algorithms.

20th Century Historical Context

The 20th century changed the world in unprecedented ways. The World Wars sparked tension between countries and led to the creation of atomic bombs, the Cold War led to the Space Race and the creation of space-based rockets, and the World Wide Web was created. These advancements have played a significant role in citizens' lives and shaped the 21st century into what it is today.

Global Conflicts & Wars

  • World War I and II
  • Cold War era
  • Regional conflicts
  • Decolonization movements

Technological Advancements

  • Aviation and space exploration
  • Digital revolution
  • Medical breakthroughs
  • Communication technologies

Political Transformations

  • Rise and fall of empires
  • Democratic expansions
  • International organizations
  • Geopolitical shifts

Social & Cultural Changes

  • Civil rights movements
  • Women's suffrage
  • Educational expansion
  • Cultural globalization

Data Characteristics

Period Coverage: 1900-1999

Content Type: Historical events and developments

Source: Wikipedia curated content

Format: Organized text with thematic headings

Phase 1: Data Collection

Project Status: COMPLETED

Phase 1 Deliverables:

  • Web scraping implementation
  • Wikipedia content extraction
  • Data storage and verification
  • Version control with GitHub

Technical Details:

Source: Wikipedia - Key Events of the 20th Century

Content: 66,144 characters, 108 sections

Method: Robust HTML parsing with fallback handling

Output: UTF-8 encoded text file

Python
BeautifulSoup
Data Storage
GitHub

Phase 2: Text Mining & NLP Analysis

A comprehensive text mining and natural language processing project analyzing key events of the 20th century using Python's NLTK and TextBlob libraries.

Project Overview

This project performs sophisticated text analysis on historical data from the 20th century, including:

  • Text Preprocessing & Tokenization
  • Frequency Distribution Analysis
  • Stop Word Removal
  • Part-of-Speech Tagging
  • Geopolitical Entity Recognition
  • Data Visualization

Technologies Used

Python 3.8+
NLTK
TextBlob
Pandas
Matplotlib & Seaborn
Jupyter Notebook

Phase 3: Network Analysis

Advanced network analysis of 20th century international relationships using graph theory and community detection algorithms.

Network Analysis Features:

  • Country Relationship Networks: Constructed from historical text data
  • Interactive Visualizations: Dynamic HTML graphs using Pyvis
  • Community Detection: Leiden algorithm for geopolitical bloc identification
  • Centrality Analysis:
    • Degree Centrality: Most connected countries
    • Closeness Centrality: Most centrally positioned nations
    • Betweenness Centrality: Key bridge countries between blocs
  • Enhanced Visualizations: Annotated plots with professional styling

Key Findings:

  • Identified distinct geopolitical communities reflecting Cold War alliances
  • Discovered key broker nations that mediated between power blocs
  • Quantified the central role of superpowers in international relations
  • Revealed network structures aligning with historical geopolitical realities

Technologies Used:

NetworkX
Pyvis
CDLib
Pandas
Matplotlib & Seaborn
89
Countries Analyzed
214
Relationships Mapped
5
Geopolitical Communities
0.078
Network Density

My Role & Process:

  • Network Construction: Built country relationship networks from historical text data using NetworkX
  • Community Detection: Applied Leiden algorithm to identify geopolitical blocs and alliances
  • Centrality Analysis: Calculated degree, closeness, and betweenness centrality measures
  • Interactive Visualization: Created dynamic network graphs using Pyvis for exploration

Network Analysis & Centrality Measures

Degree Centrality Analysis

Degree Centrality

Measures direct connections - identifies the most connected countries in the network.

Key Finding: Japan and Soviet Union show highest degree centrality, indicating extensive diplomatic relationships.

Closeness Centrality Analysis

Closeness Centrality

Identifies countries that can quickly reach others - measures strategic positioning.

Key Finding: Japan and Soviet Union again lead, showing optimal network positioning for influence spread.

Betweenness Centrality Analysis

Betweenness Centrality

Reveals bridge countries that control flow between different network segments.

Key Finding: Soviet Union dominates as key broker, controlling information flow between geopolitical blocs.

Comparative Centrality Analysis

Comparative Analysis

Shows how countries rank across different centrality measures.

Key Finding: Superpowers maintain consistent high rankings across all centrality types, confirming their dominant network positions.

Strategic Network Insights

  • Superpower Dominance: Soviet Union and Japan consistently rank highest across all centrality measures, confirming their pivotal roles in 20th century geopolitics.
  • Bridge Nations: High betweenness scores reveal countries that served as crucial intermediaries between different geopolitical blocs during conflicts and negotiations.
  • Community Structure: Network partitions align with historical alliances (Western Bloc, Eastern Bloc, Non-Aligned Movement), validating the quantitative approach.
  • Influence Pathways: Closeness centrality identifies nations best positioned to rapidly spread influence or information through the international network.

Text Mining & Linguistic Analysis

Top 20 Mentioned Countries in 20th Century

Country Mentions Analysis

Analysis of most frequently mentioned countries in 20th century historical text.

Key Finding: Germany and Japan lead with highest mentions, reflecting their central roles in major 20th century conflicts.

Part-of-Speech Tag Frequency

Linguistic Patterns

Frequency analysis of part-of-speech tags in historical text.

Key Finding: Nouns dominate historical discourse, with singular nouns (N) being most frequent, indicating focus on specific entities and concepts.

Top Nouns, Verbs and Adjectives

Word Category Analysis

Most frequent nouns, verbs, and adjectives in 20th century historical text.

Key Finding: Military and geopolitical terms dominate, with words like "war", "became", and "soviet" appearing most frequently.

Key Findings & Insights:

89
Countries Mapped
In network analysis
5
Geopolitical Communities
Identified via Leiden algorithm
0.566
Max Closeness
Japan & Soviet Union

Strategic Impact:

  • Quantified international power dynamics using network science methodologies
  • Identified key broker nations that mediated between geopolitical blocs
  • Revealed community structures matching historical alliance patterns
  • Provided framework for predictive analysis of international relations
  • Demonstrated application of graph theory to historical analysis

Tools & Technologies:

Python
NetworkX, Pyvis, CDLib
Network Science
Graph Theory Analysis
Community Detection
Leiden Algorithm
Interactive Viz
Pyvis Networks
Back to Portfolio View Code