I have recently exchanged a couple of emails with Jakob Hausladen, who has written a fascinating and enjoyable article on lute music, as part of a course on the philosophy of data science. The article draws on some of the techniques discussed on this site (including composers’ dates and nationalities, text analysis, publication network analysis), and also includes a section looking at a comparison of composers’ styles based on MIDI data. The article explains things well (including the risks and caveats), and makes excellent use of interactive graphics and sounds.

Like any good research, it raises lots of interesting questions for further investigation (some of which I have added to my list!)

I would recommend it for anybody interested in the use of statistics / data science to study music history. Or in lute music!