Doug Selent has worked with Neil Heffernan and the ASSISTments online tutoring system. His research has primarily been focused on several aspects related to large scale experimentation in online tutoring systems. He has developed a machine-learning algorithm to learn the incorrect processes students make from the students answer and the inputs of the problem and conducted several randomized controlled experiments in the ASSISTments system using this information to help students learn from their mistakes.
Doug also helped develop the Assessment of Learning Infrastructure (ALI), which is a system that automatically reports results of experiments run inside the ASSIStments system. He has developed a system called PeerASSIST. PeerASSIST is a system integrated into ASSISTments. PeerASSIST is a system which crowdsources student work and uses customized multi-objective multi-armed bandit (MO-MAB) algorithms to redistributed that work to peers in need of assistance. This system is currently being used by 60 teachers and 600 students who have generated over 300,000 instances of which over 2,000 have been redistributed to students in need of assistance.
Doug likes to teach and has taught several graduate Computer Science courses at Rivier University (Algorithms, Java Programming, Information Technology, and Object-Oriented Design) and has also programmed over a half million lines of code.