What does AskAway sound like?

The Access 2015 conference (September 8-11) included a hackfest session led by York University's William Denton, who taught attendees how to use Sonic Pi to turn data into sound. The idea is to explore different ways of representing data, which may tell you things about your data that you hadn't realized. With some rather basic Ruby coding (which the software helps you with), you can take a data set, read it into the program, and generate music. (Bill provides instructions here.) We had to pick some data to work with, and what data is more readily available and interesting than AskAway usage statistics? So I took the Form Fields data from January 2015, which includes a record for each question, and read it into the software thus:

  • Arranged questions in order of date and time entered
  • Assigned each patron's institution to one of the five AskAway tiers
  • Assigned each tier a note
  • Assigned each note a length based on the length of the chat session


Each chat session is a note, and while the method is pretty simple and it's more for fun than revelatory, it might tell you a bit more about how AskAway sessions go from month to month. A long, slower piece indicates a lot of long, in-depth sessions. A quick and lively piece indicates a lot of brief sessions. A wide variety of notes represents a variety of institution types, while a note that predominates tells us that a lot of questions are generated by one particular tier. Ideas to make something like this more useful (or sound nicer) are welcome, and I might do more with it eventually depending on how much  time I have (which is none at the moment). Perhaps eventually every statistics spreadsheet will be accompanied by an alternate mp3 option. - Brandon