Movie Dialogues Analysis

This project explores the Cornell Movie-Dialogs Corpus, combining network science and text analysis to study interactions and relationships in movie dialogues. It is part of exam project of the 02805 Social Graphs and Interactions course at DTU.

What This Project Does

• Network Analysis: Constructs interaction networks from movie dialogues, where nodes represent characters and edges represent their interactions.
• Text Analysis: Investigates the sentiment and semantics of movie dialogues to uncover emotional and thematic patterns.
• Integration: Combines network insights with text analysis to provide a deeper understanding of storytelling dynamics.

Goals of the Project

• Identify central characters and their roles in the networks.
• Explore how sentiment shapes relationships in dialogues.
• Analyze thematic clusters and patterns in movie conversations.

About the Dataset

The Cornell Movie-Dialogs Corpus contains:

• Over 300,000 lines of dialogue.
• Metadata about movies, characters, and conversations.

Potential Applications

• Understanding narrative structures in movies.
• Analyzing emotional and thematic trends in storytelling.
• Visualizing character dynamics and relationships.

This project provides a foundation for exploring the intersection of social networks and language in cinematic storytelling.

Paper with rich features for the movie dialogues https://arxiv.org/pdf/2303.16618