Fake Tweet Detection System

Python Flask app in Docker which detects whether a tweet is factually correct or inaccurate. This app is an interactive dashboard that displays the flow of truths and lies that stream through the platform each day, using NLP to flag tweets that may contain misinformation.

POST: http://34.123.199.55/predict_tweet with a json {"tweet": "the tweet text goes here"} and will receive a json {"result":"fake"}

Build application

Build the Docker image manually by cloning the Git repo.

$ sudo docker build -t gcr.io/fake-tweet-detection/fake-tweet-backend:v1 ./

Run the container

Create a container from the image.

$ sudo docker run --rm -p 8080:8080 gcr.io/fake-tweet-detection/fake-tweet-backend:v1

Push the container to cloud

Push the container to GCP:

$ sudo docker push gcr.io/fake-tweet-detection/fake-tweet-backend:v1

Now visit http://localhost:8080