As the final project of the Coursera / Johns Hopkins data science series that I’ve just completed, I made this neat app.

See my code here.

It’s similar to the predictive keyboards you can buy for your smart phone, but I added in a fun little babbling tool.

The Coursera data science specialization is centered around the R statistical programming language and consists of ten classes that cover programming in R, statistics and probability, processing and analyzing data, machine learning, etc. I jumped into it in its first iteration, so I was somewhat of a guinea pig, but the series was very well-organized and entirely fascinating, except for the last class, which was essentially untaught and a bit overwhelming. It involved learning several topics from a huge field (Natural Language Processing) of which most of the students knew nothing, and then learning how to process a large amount of data (just over 100 million words), which my macbook pro (with 8 GB of RAM) proved incapable of doing (I ended up renting an Amazon EC2 server with 16 GB of RAM), and then writing a prediction algorithm and web application. Anyway, I did it! It’s been done before, but now I’ve done it too.