Google Cloud Streaming Twitter Example

This examples shows how to listen to a Twitter hashtag (via the streaming API), process the tweet's sentiment and store the results all within Google Cloud Platform.

When a tweet is received, the application will run a sentiment analysis using Google's Natural Language Processing and then store the results within BigQuery.

Prerequisites

• git installed - https://git-scm.com/book/en/v2/Getting-Started-Installing-Git
• node installed - https://nodejs.org/en/
• If running within Google Cloud Compute Engine make sure to enable API scopes when creating the VM

Setup

1. git clone https://github.com/joeint/gcp_twitter_example.git
2. npm install
3. Generate Twitter keys from https://apps.twitter.com/
   • input keys into the .env file
4. Create the Big Query dataset and table
   • bq mk twitter
   • bq mk --schema HashTag:STRING,Tweet:STRING,SentimentScore:FLOAT,SentimentMagnitude:FLOAT,InsertDate:STRING -t twitter.twitter_stream
5. Update .env file with the GCP project id that contains the BQ Twitter table
6. (optional) update the hashtag in app.js

Running the application

You can run the application as on a virtual machine or within app engine

VM Option

node app.js

App Engine

gcloud app deploy