This project is an AI-powered tool designed to extract and analyze contract specifications from PDF documents, with a focus on construction contracts. It leverages natural language processing (NLP) techniques to identify key information such as materials, tests, standards, and definitions.
🌟 Key Features
Extraction of Materials, Tests, and Standards: Automatically extracts relevant information from construction contract PDFs.
Intelligent Q&A System: Enables users to query the document (e.g., "What are the requirements for M25 concrete?") with answers derived from extracted data and full text..
Support for Large Documents: Optimized for processing PDFs with 100+ pages.
OCR Fallback: Employs Tesseract OCR for scanned documents when text extraction fails.
🛠️ Technologies Used
Python
pdfplumber: For PDF text and table extraction.
spaCy: For NLP tasks and entity recognition.
Transformers (BERT, DistilBERT): For named entity recognition (NER) and question-answering.
Gradio: For the interactive web interface.
FAISS: For efficient similarity search in the Q&A system.
ReportLab: For generating PDF reports.
Tesseract OCR: For optical character recognition.
Sentence Transformers: For semantic search in Q&A.
👥 Team Members
Gaurav Tarate | Shubham Palve | Atharav Pawar
📄 License
This project is licensed under the MIT License. See the LICENSE file for details.