Inspired by this post on recommendation systems and this one on classification I’ve been playing around with some of the large datasets on my different projects.

For the Chicago Maroon we have a giant database of articles. I wanted to see if it would be possible to classify articles into a section “News”, “Viewpoints”, “Voices”, “Sports” based on the text of their lede.

It’s actually pretty easy, check out the code below for how.

I’ve also been playing with making a recommendation system for colleges over at CaptainU. It’s working pretty well, but I can’t share.

Here’s the code for the lede classifier: