“Why we need new cats”: The challenges of making a TED Talk title generator

As part of a project for a Natural Language Processing class last spring, I wrote a program that generated names of fake TED Talks.  The goal was to get the generator to produce titles that made sense using only the information contained  in the pre-existing titles available online; I didn’t want to write a bunch […]

Read More

Parsing partisanship

Computational Linguistics doing cool stuff: The folks at FiveThirtyEight wrote a program to predict the political leanings of economists, using nominally scientific, non-partisan academic papers as a learning corpus. Apparently certain phrases can reveal the political beliefs of their authors, if you know what to look for. Read more about it here!

Read More

Words for the Dumpster

Here’s Timothy Egan on what words and phrases he never wants to hear again.  I have to agree with him on most of these.  I would also like to add “ironic” and “unique,” which don’t really mean anything anymore.

Read More
Skip to toolbar