The first recommendation engine is Common Learner Recommendation. So I'm just going to turn this on, and this is actually the easiest to understand because what we're doing is we don't need administrator to do anything other than enable this engine, this recommendation engine. And then we're just going to use people's, all the learners' learning history to cross-recommend so that if we notice that Ashley took a toothpaste course and then we also notice a bunch other learners who also took the toothpaste course have taken toothbrush course, we're going to recommend Ashley the toothbrush course simply because the actual learner themselves is using the system and their behavior is essentially telling you what course should recommend. This works for some courses, not others, which is why we have other recommendation engine that you can add to the mix. When it does not work well is when you have a few very popular courses that everybody takes, then you end up having recommendation that is identical to everybody regardless what they have taken because a few popular courses essentially overwhelm other courses that would have been recommended because they're just too popular, right? But when this is turned on, what you can do is we can go to the courses and then let's take a look at course, this Southern Rise and Fall of Telehealth. When I click on it, I can click on related product. You will see here is all the other courses in the system that have at least two users that has taken it so that you can see that, oh, I'm sorry, at least one user to say, hey, there's at least one user who have taken this course and this course right here. And there are three users who have taken this course and have taken this course. So obviously the system I'm using here is just a demo site that doesn't have a whole lot of traffic. But if you have a very busy site, you will definitely see a lot more courses under related product where you're going to see a long list of courses that could be on the recommendation list sorted by how many common learners they have. Notice that we do track common learners in the last 12 months and also the common learner lifetime. And the idea here is that if a course have a lot of common learners that is more recent in the last 12 months, that is more indicative of those two courses are somewhere related. But the tiebreaker is going to be lifetime so that if there's multiple courses that have the same number of common learners within the last 12 months, then we're going to prioritize the one that has more common learner in the lifetime as a tiebreaker. So that's the logic of how the common learner recommendation works. Next video, we're going to show exactly what the learner sees based on what they have done. Thank you.

recommendation engine

course suggestions

learner history

popular courses

diverse offerings