Decide on what the package we're creating will do

[lwjohnst86]

As well as what type of data to include. Right now it uses books from the Gutenberg Project. So we could just keep using that as a simple example and have the project work use more classic numeric data? Tho, nice thing with the book/text data is that we can do some ML/NLP stuff on it, which might be interesting to participants.