Temporal Knowledge Graph of India-Pakistan Relations

This project constructs a Temporal Knowledge Graph to analyze the evolution of relations between India and Pakistan. By parsing historical text data, it extracts entities (people, places, events) and models time-varying relationships to represent how interactions have changed over time.

Project Overview

This Python-based model uses NetworkX and spaCy to build a knowledge graph where:

• Nodes represent entities such as people, events, and places.
• Edges depict temporal relationships like conflicts or alliances.
• Querying capabilities allow users to explore historical conflicts and alliances over time.

Key Features

1. Entity Extraction: Automatically extracts people, places, and events from historical_data.csv.
2. Temporal Relationship Modeling: Uses NetworkX to model relationships that evolve over time.
3. Timeline Summary Generation: Summarizes major events, conflicts, and alliances for each year in the dataset.
4. Interactive Querying: Enables questions like:
   • "Which countries were in conflict during 1965?"
   • "How did India-Pakistan relations evolve from 1947 to 1999?"

Requirements

• Python 3.x
• NetworkX for graph processing
• pandas for data handling
• spaCy for entity extraction

To install dependencies, run:

pip install networkx pandas spacy
python -m spacy download en_core_web_sm

Code Overview

The project contains two main scripts:

1. file1.py: Creates the knowledge graph and allows querying of relationships.
2. file2.py: Adds functionality for generating a timeline summary of events, classifying relationships as conflicts or alliances.

file1.py - Knowledge Graph Creation and Querying

• load_data(): Loads historical data from historical_data.csv.
• extract_entities(): Extracts entities (people, places, events) using spaCy.
• extract_temporal_relationships(): Identifies relationships between entities by year.
• create_knowledge_graph(): Builds a graph with nodes for each entity and edges for relationships over time.
• query_graph(): Queries the knowledge graph for specific relationships, such as conflicts or alliances in a given year.

file2.py - Timeline Summary Generation

• timeline_summary(): Produces a timeline of events for each year, categorizing relationships as conflicts or alliances based on context.
• print_timeline_summary(): Displays the timeline in a readable format.

Usage

1. Run file1.py to create the knowledge graph:

   python file1.py

   • Includes examples of querying for conflicts in specific years and the evolution of relations between entities.

2. Run file2.py to view a timeline summary:

   python file2.py

   • This script outputs major events by year, noting conflicts and alliances.

Example Queries

• Conflict in 1965: Returns relationships categorized as conflicts for the year 1965.
• Evolution of Relations: Shows how the relationship between entities (e.g., India and Pakistan) evolved over time.

Example Timeline Summary Output

**1947**
- India and Pakistan: Conflict

**1965**
- India and Pakistan: Conflict

**1971**
- India and Bangladesh: Alliance

**1999**
- India and Pakistan: Conflict

Contributing

Contributions are welcome! You can enhance entity extraction, refine relationship modeling, or expand query capabilities.

License

This project is licensed under the MIT License.