Challenge 1b: Persona-Based Document Analysis

Overview

This solution analyzes PDF document collections based on persona and job descriptions to extract relevant sections and perform subsection analysis.

Tech Stack

• Python 3.10: Core programming language
• PyMuPDF (fitz): PDF text and image extraction
• spaCy: Natural language processing and keyword extraction
• Tesseract OCR: Image text extraction with multilingual support
• scikit-learn: TF-IDF vectorization for keyword analysis
• langdetect: Language detection for multilingual support
• concurrent.futures: Parallel processing for performance optimization
• pathlib: Cross-platform file path handling
• json: Input/output formatting

Features

• Multilingual Support: Hindi, English, Japanese, Arabic, Chinese, Russian, French
• Persona Analysis: Extracts relevant sections based on persona and job requirements
• Keyword Extraction: Uses TF-IDF and spaCy for intelligent keyword analysis
• Image OCR: Extracts text from images within PDFs
• Performance Optimized: Fast processing with parallel execution

Input

• JSON file with persona, job description, and document list
• PDF files referenced in the input JSON

Output

• JSON file with extracted sections and subsection analysis
• Relevance scoring for each section
• Keyword alignment with persona and job requirements

Usage

python r1b_persona_analyzer.py input.json output.json

Performance

• Execution time: ≤ 10 seconds for 50-page PDF
• Model size: ≤ 200MB
• Network: No internet access required
• Runtime: CPU with 8 cores, 16GB RAM