This application is available on https://tweet-dashboard.mybluemix.net

The purpose is to analyze tweets doing sentiment analysis in real time.

ToDo:

• Create a view to retrieve number of tweets by hour.
• Refactor cloudant.js to use more than one DB at a time.
• Download CSV writing to disk instead of memory.
• Change stuff from router to controller.
• Make chart update automatically.
• Create a model for tweets.
• Create tabs on website by candidate.
• Create some way of classifying tweets by candidate.
• Clean old tweets from front-end in order to prevent memory leak.
• Put the view code on github
• Create CSV script to run as child_process
• Filter URLs and twitter usernames, as they tend to impair duplicate filter
• Find some way of making a middleware on cloudant.js that checks if the database exists before making access to cloudant object, like a middleware.
• Use gulp.js to differentiate between dev test and prod envs
• [ ]

1. npm install
2. DEBUG=app:* npm start