Project 5: Shakespeare Sentiment Analysis. (code available on Github)
Scope: Use the existing Node.js / Express / sentiment scaffolding from Project 4 to perform sentiment analysis on the top six speakers in a selected Shakespeare play.
You can read more about the sentiment module and sentiment analysis in the write-up for Project 4.
What I Learned:
- Server-side parsing: I thought this would be a straightforward snap, since I already had the sentiment analysis logic (from Project 4) and the Shakespeare-play-parsing logic (from Project 1). It was mostly straightforward, except that I quickly realized there are certain things (simple jQuery and DOM functionality) I’d taken for granted on the client side that weren’t available to me on the server side. I wound up having to use the jsdom module and the string.js module to help me out.
What I Left:
- I could’ve jazzed it up a bit more with some “Loading” signifiers (beyond just disabling the buttons while it loaded), and possibly added some more plays.
- I also could have added some emoticon-type images to indicate whether the character was positive, negative, or neutral. At present, a numerical score is given with no indication of where that number falls on the “happy/sad/neutral” spectrum.