This is a self-contained service utilizing the NLTK for sentiment classification.
Environments:
- development
- production
Usage:
cd docker
docker build -t sentiment_service:1.0 .
docker run -d -P --name sentiment_service -e ENV=development -v /path/on/host:/application sentiment_service
[OR]
docker run -d -P --name sentiment_service -e ENV=production sentiment_service
Todo:
- Python 3 (DONE)
- flask (DONE)
- uWSGI (DONE)
- nginx
- NLTK
- set up features
- generate training corpus
- test analysis
- adjust features