Legal Document Analysis - Unleashing the Power of AI on Legal Text

Workflow

Here's a visual representation of the workflow:

graph LR
A[Input PDF Document] --> B(Text Extraction)
B --> C{OCR}
C --> D(Text Preprocessing)
D --> E{Summarization Flan-T5}
D --> F{Named Entity Recognition BERT}
E --> G[Summary Output]
F --> H[Named Entities Output]

Loading

Live Demo

Try out the Legal Document Summarization model with our interactive demo hosted on Hugging Face Spaces:

Dive into the World of Automated Legal Insights!

This repository empowers you to effortlessly analyze legal documents using the cutting-edge capabilities of Natural Language Processing (NLP). Say goodbye to tedious manual reviews and unlock valuable insights hidden within complex legal text.

Imagine:

• Effortlessly summarizing lengthy legal documents to grasp key information in minutes.
• Instantly identifying crucial entities like individuals, organizations, and locations mentioned in the text.
• Extracting key information to streamline your legal research and decision-making process.

This repository provides the tools you need to:

• Summarize Legal Documents: Condense lengthy legal texts into concise summaries using a fine-tuned Flan-T5 model.
• Extract Named Entities: Identify and categorize key entities like people, organizations, and locations with a powerful BERT-based NER model.
• Analyze Text from PDFs: Extract text from PDF documents, including OCR capabilities for scanned documents.

Key Features:

• State-of-the-art NLP Models: Leverages pre-trained and fine-tuned models for superior performance.
• Easy-to-Use Interface: Clear and concise code makes it simple to integrate into your workflow.
• Customizable: Adjust parameters to fine-tune the analysis to your specific needs.
• Comprehensive Analysis: Provides both summaries and detailed entity extraction for a holistic understanding.

Getting Started:

1. Clone the Repository:

   git clone https://github.com/shresthasingh1501/legal_document_analysis.git

2. Install Dependencies:

   pip install -r requirements.txt

3. Explore the Code:

   • Fine-tuning Code:

     • BERT for NER: Find the code for fine-tuning the BERT model for Named Entity Recognition in the notebooks/Entity_ExtractionUsing_BERT_&_PyTorch_finetune.ipynb notebook.
     • Flan-T5 for Summarization: Explore the code for fine-tuning the Flan-T5 model for summarization in the notebooks/FlanT5-Summerizer.ipynb notebook.

   • Core Analysis Scripts:

     • analysis.py: This script provides the core functionality for processing PDF documents, summarizing text, and extracting named entities.
     • bert-ocr.py: This script enhances text extraction by incorporating OCR capabilities for scanned PDF documents.

4. Test with Sample PDFs:

   • Place your sample PDF documents in the samplepdfs directory.
   • Modify the scripts to point to the correct directory and run them to see the magic in action!

Where to Find What:

• code/: Contains the Python scripts for analysis.
• notebooks/: Jupyter notebooks for fine-tuning models.
• readme.md: This informative guide!
• requirements.txt: List of project dependencies.
• samplepdfs/: A directory to store your sample PDF files.
• streamlit-ui.py: (Optional) A Streamlit-based UI for interactive analysis (implementation in progress).

Unleash the Power of AI on Your Legal Documents Today!

This repository provides a powerful toolkit for automating legal document analysis. Start exploring and discover the insights waiting to be uncovered!

Contributions are welcome! Feel free to open issues, suggest improvements, or submit pull requests. Let's build a better future for legal analysis together.

Disclaimer: This repository is for educational and research purposes. It is not intended to provide legal advice.