Taranis AI Sentiment Analysis

This project integrates sentiment analysis into Taranis AI, allowing for the classification of news items as positive, negative, or neutral using transformer models. The API intelligently chooses the appropriate model based on the text length, utilizing XLM-RoBERTa for shorter texts and Longformer for longer texts.

Features

• Multi-language Support with XLM-RoBERTa and Longformer.
• Pre-trained Models for high-accuracy sentiment classification.
• Flask API for external calls to the sentiment analysis.

Setup

Prerequisites

• Python 3.10+

Installation

To install dependencies:

uv venv
uv sync

Running the API

To start the Flask API:

flask run
# or
granian run

# or
docker run -p 5000:5000 ghcr.io/taranis-ai/taranis-sentiment-bot:latest

Example API Call

To test the API with a POST request, use curl:

curl -X POST http://127.0.0.1:5000/ \
  -H "Content-Type: application/json" \
  -d '{"text": "This is an example sentence to analyze sentiment."}'

Example Response

{
  "sentiment": {
    "label": "POSITIVE",
    "score": 0.94
  }
}

License

EUROPEAN UNION PUBLIC LICENCE v. 1.2