encoder-010/infotact-project1-grievance-nlp

An AI-powered system for automated analysis of citizen complaints: it classifies which government department should handle a complaint, detects sentiment (Critical/Negative/Neutral/Positive), assigns an urgency score (0–100), and recommends priority actions. Intended users include municipal/government grievance teams, civic-tech developers, and evaluators of NLP solutions. Key features: - Models: TF-IDF vectorization with Random Forest for department classification and SVM for sentiment. - Tech stack: Python, NLTK, scikit-learn, FastAPI (REST endpoints: GET /, GET /health, POST /analyze). - Performance: department accuracy 53.33%, sentiment 66.11%, macro F1 53.33%. - Run: pip install -r requirements.txt; uvicorn main:app --reload. - Organized week-wise development and documented commits.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.