For my final project for Nature of Code Intelligence and Learning, I am going to go back to my first project and use it to convey a data set.

My initial project was the three dimensional binary search tree in three.js. The sketch generates 50 spheres of random sizes and sorts them. As a reminder, when run, it generates something that looks like this:

At the time I created it, I tried to find data that might sort well in this format, but my couple of tests didn’t work out well. So, I’m going back to that idea, and I’ve decided to use personal text messages as the data set.

The Data Set

I am going to use iMessages sent between me and my partner for the past four weeks as my data set. I have already downloaded and sorted the data, and I have done character and word counts on each of the messages. (They were in an SQLite database in ~/Library/Messages/chat.db.)

I thought this would be an interesting dataset to work with because I haven’t done much work at ITP with personal data. It’s nice to have a first time for everything, no?

The Representation

Each day will get its own binary tree representation. Each sphere will represent one message and will be sized by either the amount of characters or words in the message (tbd per testing). Each of us will have our own color.

It’s a simple idea, but I’m excited to see what it will look like.