As part of my dissertation, I plan to code the courses in my data set according to what subject they cover. I have the option of using a subject prefix (e.g. "ENG203" is an English course) but if I develop a model using that subject designation as an indicator, it won't be very reusable.
So after doing some research, I've decided to adopt the UNESCO International Standard Classification of Education: Fields of Education and Training 2013 (ISCED-F 2013). You can see it here:
I've taken this standard and implemented it as a Competency Framework in Moodle. You can get the framework here: https://moodle.net/mod/data/view.php?d=12&rid=861
Next, I'll go through my course site and attach the relevant "competency" (subject code) to each course. Then I can create an indicator based on these known competencies. I've given each one a code that I think should be fairly distinct from other likely competency codes.
What this means for everyone else: once I've got the indicator working, I can publish that as a plugin and anyone who wants to add that indicator to an existing (or new) model can install it, import the competency framework, and tag courses on the site. I'll let you know if including this indicator seems to significantly improve accuracy of predictions in my data.