Matches and Mismatches

Awhile back I had an exchange with one of the OkCupid founders about if there were any ideas from matching algorithms used on dating sites that could translate to models for improving matches of students to teachers/education.  Coincidently, this Atlantic article came out a few months after our conversation which more deeply describes the higher education mismatch issue.  Dating sites have been trying to crack the nut of a more human matching method for awhile.  Similarly music sites like Spotify are also trying different tactics for sourcing personal preference data to better match people to people and/or things.  With the recent emergence of online university courses like Udacity and Coursera the analogies have crossed my mind again.  While topic and level are one generic cut (like gender and age is in dating), I wonder how data about other qualities, like delivery style, could help the growing number of online learners better find the class they’ll stick with.