Note: This is a post that I started some time ago and have had in my todo list to finish for...maybe a year now? Apologies for the delay!

We've argued that more entropy in a survey is better for detecting bad actors. The argument goes like this: A survey of 5 yes/no questions has (ignoring breakoff) 32 possible unique answers. The maximum entropy of this survey is

$$-5\bigl(\frac{1}{2} \log_2(\frac{1}{2}) + \frac{1}{2} \log_2(\frac{1}{2}) \bigr) = -5\bigl(-\frac{1}{2} - \frac{1}{2}) = 5$$.

I was just looking at some old crappy code I wrote four years ago, to show Molly that one should have no shame when posting code online. While looking at my own HMM code, I ran across this code. I have no idea what it does. I just remember wanting to use Perl for string munging crap. Yikes!

In response to a Keanu moment I recently had, Arjun recommended I read On the expressive power of programming languages. The paper is divided into roughly two parts: in the first part, Felleisen sets up the formal framework for describing the expressiveness of programming languages. In the second part, he illustrates the repercussions with Scheme.

Participation and Contribution in Crowdsourced Surveys, a recent PLOSOne article, discusses some interesting approaches to crowdsourced surveys. Not only are the answers crowdsourced, but the questions themselves are also crowdsourced. The surveys are seeded with a small number of questions and later augmented with questions supplied by respondents. These questions are curated by hand and presented to respondents in a random order.

Woah, it's been a while since I actually made a blog post. I'm going to to try to make a blog post every day until I head off for my Facebook internship in June (super psyched!). I'm putting it in writing, so that's how you know I'm definitely going to break my word...

I owe the above title to Bryan Ford, who listened to me raging against Machine Learning this afternoon. Bryan is visiting for our first systems lunch in a while. It's been a long hiring season...