README.md

metadata

title: Sentiment Analysis API
emoji: 😊
colorFrom: blue
colorTo: green
sdk: docker
pinned: false

Sentiment Analysis API

A FastAPI-based sentiment analysis service using ensemble machine learning models (KNN, Random Forest, Extra Trees).

Features

• Fast predictions using pre-trained ML models from HuggingFace Model Hub
• RESTful API with automatic documentation
• CORS enabled for web frontend integration
• Ensemble learning for improved accuracy

API Endpoints

GET /

Health check and API information

GET /status

Returns model status and readiness

POST /predict

Analyzes sentiment of input text

Request body:

{
  "text": "Your text here"
}

Response:

{
  "predicted_sentiment": "positive",
  "input_text": "Your text here"
}

Models

This API downloads models from HuggingFace Model Hub: anis80/anisproject

• Label Encoder: Encodes sentiment labels
• TF-IDF Vectorizer: Converts text to numerical features
• Voting Classifier: Ensemble of KNN, Random Forest, and Extra Trees

Documentation

Interactive API documentation available at /docs (Swagger UI)