ulrichglass / ironhack-final-project

ML Pipeline for sentiment analysis on Twitter.

Ironhack Final Project

Public files for my final project completed at Ironhack Data Analytics Bootcamp in December 2020.

Overview

The idea was to build a complete pipeline:

• fetching tweets directed to a @AppleSupport via API (excluding retweets and replies);
• storing them on a MySQL Database on Amazon Relational Database Service (RDS);
• sending the tweets to Microsoft Azure Cognitive Services for sentiment analysis and opinion mining via API;
• updating and storing the enriched data on Amazon Relational Database Service (RDS);
• automate the fetching with an AWS Lambda function and sending notifications if a certain threshold is reached via Amazon;
• the results are imported into Tableau Public (no direct link to SQL database possible in this version);
• presenting the results on a website hosted on AWS Elastic Beanstalk via Flask.

Pipeline

Website

http://nlp-twitter-test-env.eba-cw2p6wnm.us-east-2.elasticbeanstalk.com
(not available all the time due to AWS free tier limits)

Home

Testing

New Tweets

Archive