StyleSynth

🎬 Hey there, Movie Buff! 🍿

Do you use Letterboxd to track and review all the amazing films you've watched? 🎥✨

Ever found yourself wondering what to do with all that user data you've generated? 🤔💭

How about spicing things up by generating personalized movie reviews that reflect your unique writing style, all thanks to cutting-edge LLM models? 🖋️💡

Got an OpenAI API key and a few spare coins for tokens? 💰🔑

Well, you’re in luck! 🎉

Introducing StyleSynth – your new best friend in the world of movie reviews! 🎊

This Python tool analyzes your personal Letterboxd review style and whips up LLM-powered movie reviews that are tailored to sound just like you! 🤖🎞️

This project is not just a tool; it’s a proof of concept showcasing the power of NLP and LLMs to create something fun and useful for all you movie enthusiasts out there! 🌟

Stay tuned for more features and improvements! 🚀

Features

• Analyzes Letterboxd reviews to understand writing style
• Identifies common patterns in reviews
• Uses OpenAI's GPT models for personalized review generation
• Vector similarity search for finding similar movies that you've watched

Quick Start

Prerequisites

• Python 3.11 or higher
• Poetry (Python dependency management tool)
• OpenAI API key
• Download Letterboxd user data and store in data/letterboxd (reviews.csv and watched.csv)

Installation

1. Install Poetry (if not already installed):

   pip install poetry

2. Clone and Set Up the Project:

   # Clone the repository
   git clone https://github.com/xXJohamXx/StyleSynth.git
   cd stylesynth
   
   # Install dependencies using Poetry
   poetry install

3. Set Up Environment Variables:

   # Option 1: Create or overwrite .env file
   echo "OPENAI_API_KEY=your-api-key-here" > .env
   
   # Option 2: Manually create .env file
   # Create a new file named .env and add your API key:
   OPENAI_API_KEY=your-api-key-here

   Replace your-api-key-here with your actual OpenAI API key.

Simple Usage Example

1. Check out the example script demo_review_generator.py which demonstrates basic usage of the library.

2. Run the script using Poetry:

   poetry run python demo_review_generator.py

Note: Make sure you have set up your OpenAI API key in the .env file before running the example.

Contributions are welcome! Please feel free to submit a Pull Request.

System Architecture

graph TB
    %% Style definitions
    classDef input fill:#fff3e0,stroke:#e65100
    classDef process fill:#f3e5f5,stroke:#4a148c,stroke-width:2px
    classDef llm fill:#d5e8d4,stroke:#82b366,stroke-width:2px
    classDef vectorb_db fill:#f8cecc,stroke:#b85450,stroke-width:2px
    classDef output fill:#bbdefb,stroke:#0d47a1 

    subgraph Input["<h3>📁 Input Files<h3>"]
        R[(Letterboxd Reviews csv)]:::input
        W[(Watched Movies csv)]:::input
        Q[(User Movie Query)]:::input
    end

    subgraph Analyzer["<h3>🔍 Learn Style Pipeline<h3>"]
        RSA["Style Analyzer<br/>(Core Process)"]:::process
        AV["Vocabulary Analysis<br/>(Async / LLM)"]:::llm
        AS["Sentence Analysis<br/>(Async / LLM)"]:::llm
        EM["Movie Embeddings<br/>(Async / LLM)"]:::vectorb_db
        PRS["Personal Style<br/>Profile"]:::output
        
        RSA --> AV & AS
        RSA --> EM
        AV & AS  --> PRS
    end

    subgraph Generator["<h3>🎬 Review Generation<h3>"]
        RG["Review Generator<br/>(Core Process)"]:::process
        FS["Similar Movies<br/>Search"]:::vectorb_db
        EQ["Embed User Movie Query <br/>(LLM)"]:::vectorb_db
        GR["Generate Review<br/>(LLM)"]:::llm
        
        RG --> EQ --> FS --> GR
    end

    subgraph Output["<h3>📝 Generated Review<h3>"]
        RT["Review Text"]:::output
        SC["Style Match<br/>Confidence"]:::output
        RE["Style Elements<br/>Used"]:::output
    end

    %% Main Flow Connections
    R & W ==> RSA
    Q ==> RG
    PRS ==> RG
    GR ==> RT & SC & RE

Loading

System Architecture Legend

Type	Description
📁	Input Files (Orange) - CSV files and user movie query
⚙️	Core Processes (Purple) - Main processing components
🤖	LLM Operations (Green) - Language model interactions
🔄	Vector Store Operations (Red) - Embedding and similarity search
📝	Output (Blue) - Generated review and analysis

License

This project is licensed under the MIT License - see the LICENSE file for details.