Jacob Munson (Dept. of Mathematical Sciences, MSU) 

4/6/2023  3:10pm

Abstract:

In this talk we will briefly review collaborative filtering and discuss its application to educational data mining. Collaborative filtering, a major paradigm of recommender systems, will be discussed in the context of rating prediction. We will draw a connection between ratings data and student grade data and discuss modeling results on historical grade data from Montana State University. We find that relatively simple baseline models, often not explored in existing literature, are effective at producing "predictions" competitive with more complicated modeling strategies. We also find tangible effects on grade dynamics following the onset of the Covid 19 pandemic.