Tuesday, September 29, 2009

0 Using recommender systems to teach math concepts

I'm sitting in my Advanced Topics in Web Technologies class with Alex Labrinidis; today's discussion is recommendation systems. For example, movie, book or music suggestions/recommendations. We are all familiar with Amazon.com recommending books that we'd like to purchase based upon our previous purchases to their home page.

Now, this is the first time I've "engaged" in learning about recommendation system, but my mind is always pondering about ways to introduce these concepts to high school students.

  1. Recommendation systems use a correlation score (eg. Pearson's correlation, Euclidean distance) to determine similar users.
  2. Problem Solving: explain the algorithm used to determine correlation (likeness). Give the students a sample data set. Ask them to determine which users have the highest correlation.
  3. We can then take it a few steps further, having the students create a "recommendation system" by hand and requiring them to determine the best method of correlation and answering some critical thinking questions.
Class is dismissed.

0 comments:

Post a Comment

Note: Only a member of this blog may post a comment.