Visual Analytics of Character Dynamics and Sentiment Evolution in Friends

This project introduces a visual analytics approach to analyzing the scripts of the iconic TV series Friends. By leveraging transcript data and advanced visualization techniques, the research uncovers hidden insights into character personas, emotional trajectories, and relational dynamics.

Key Features

• Character Persona Analysis: Identifies traits, behaviors, and emotional patterns of main characters.
• Sentiment Dynamics: Tracks emotional trajectories across episodes and seasons, highlighting pivotal moments.
• Character Interaction Evolution: Examines how relationships evolve and shape the show's narrative structure.
• Integrated Techniques: Combines sentiment analysis, network visualization, and narrative analytics to reveal key trends.

Insights

• Discover how character relationships (e.g., Ross and Rachel) evolve over time.
• Explore emotional peaks and valleys that align with dramatic moments and season finales.
• Visualize the impact of character interactions on the show's narrative structure.

Applications

This framework is not limited to Friends—it can be extended to analyze other television series, movies, or narrative-driven media, offering fresh perspectives on storytelling and character development.

Repository Contents

• /Data: Contains cleaned datasets (transcripts, sentiment labels, and metadata).
• /Notebooks: Includes Python scripts and Jupyter notebooks for sentiment analysis, network visualization, and temporal trend mapping.
• /Results: High-resolution visualizations (network graphs, heatmaps, spider web charts, etc.).
• /docs: Project paper and additional documentation.

How to Use

1. Clone the repository: git clone https://github.com/your-username/Friends-Visual-Analytics.git
2. Install dependencies: pip install -r requirements.txt
3. Run the scripts:
   • Analysis: /Notebooks/Friends_analysis.py
4. Explore the visualizations in /results and insights in the project paper.

Example Visualizations

• Network Graph of Character Interactions: Visualize relationships and sentiment intensity between characters.
• Sentiment Line Graph: Track emotional highs and lows across episodes.
• Heatmap of IMDb Ratings and Viewership: Analyze audience engagement trends.

Results Overview

Emotion Distribution

Character Analysis

Character relation

Future Work

• Extend sentiment analysis to all 10 seasons using advanced NLP models (e.g., BERT).
• Develop interactive dashboards for exploring character-specific subplots.
• Apply the framework to other TV series or narrative-driven datasets.

References

• Full Project Paper
• Dataset Source