Wanted to play with Twitter api as well as try to deploy something with docker.

What it does

Tracks the number of positive vs negative tweets that contain a particular search term/tag.

How we built it

It runs a separate Scala/Akka/JVM based dockerized backend hosted on an AWS EC2 instance orchestrated by TumTum (Docker Cloud) which is connected to the Twitter stream. The API lets you chose tags/search terms to follow. When a tweet is matched it is sent to the Azure Text Analytics API for sentiment analysis. A running total of positive vs negative tweets is kept for each search term, available on another API. API GitHub

The front end is written in Javascript using Webix and hosted on AWS S3. Webview GitHub

Challenges we ran into

Staying awake long enough to finish. Getting the Java Twitter4j library to play nicely with Scala - Scala primitives are not always the same. CORS - enabling the backing to take requests from other domains. Many many more...

Accomplishments that we're proud of

It works without crashing \m/ (so far...)

What we learned

Twitter API. Azure API. Akka Actors api. Docker/Tumtum.

What's next for Bits Please

More analytics and persistence of results. Better UI. Tried to use the domain voucher to buy for the project but's site was being less than useful. I think I have managed to buy it now, but they won't send me the details to log in.... meh.

Share this project: