Being able to quickly summarize a piece of literature is something that is quite useful before deciding whether or not you want to read it. However, we all know judging a book by its cover isn't quite the right thing to do. However, being able to quickly determine the "mood" throughout the piece would be a good place to start.
What it does
Applies a naive bayes classifier to the given document, and uses this to determine the average mood of each sentence, and from that, determines the average mood throughout the document.
How I built it
National Language ToolKit supplied the data for the classifier, and python allowed a convenient way to implement this model.
Challenges I ran into
Building the Classifier without adequate test data Getting python to play with audio files nicely
What's next for Sensing Semantics
Getting better test data