NLP Analysis of the Red Rising Book Series

An information retrieval and text mining project analyzing the first two books of Pierce Brown's Red Rising series. Uses NLP techniques to extract story themes over time and visualize character relationship networks.

📄 Report: IRTM_Assignment.pdf

What This Project Does

• Topic modeling and topic river visualization: Extracts the main themes from each section of the books and visualizes how topics evolve across chapters, revealing the narrative arc through data.
• Named entity recognition: Identifies characters, locations, and factions mentioned throughout the text using BERT-based NER.
• Coreference resolution: Links pronoun references back to their corresponding characters to build accurate interaction counts.
• Character network graphs: Constructs and visualizes relationship networks showing which characters interact most frequently, using both chapter-based and sliding-window co-occurrence methods. Visualized with Gephi.

Repository Structure

├── entity_recognition.ipynb   # NER pipeline
├── event_recognition.ipynb    # Event extraction
├── coref_bert.py              # BERT-based coreference resolution
├── divide_chapters.py         # Chapter segmentation
├── tokenizer.py               # Text preprocessing
├── web_scraping.py            # Data collection
├── result_visualisation/      # Generated figures and plots
├── coref_chapters.gephi       # Gephi network graph (chapter-based)
└── coref_windows.gephi        # Gephi network graph (window-based)