slack-sentiment

An example of running sentiment analysis on a slack channel.

Getting Started

This project required Python. Ideally 3.9 or above.

Packages are managed with pipenv.

Install dependencies

pipenv sync --dev

Run notebook

pipenv run notebook

Deploying a Model with Flask and Docker

Assuming you have created a model by running through the notebook, these instructions will show you how to take that model and expose it as a web service deployed using a docker image.

Build the image

docker build -t slack-sentiment .

Run the container

docker run -it --rm -p 5000:5000 slack-sentiment

If you want to leave the container running in the background instead use:

docker run -d -p 5000:5000 slack-sentiment

Test the model

$ curl -d "text='Hello! Have a wonderful day.'" -X POST http://localhost:5000/

That text appears to be positive :)

$ curl -d "text='Boo. This sucks!'" -X POST http://localhost:5000/

That text appears to be negative :(