Recommender Systems

Daily writing prompt
Describe a positive thing a family member has done for you.

This daily writing prompt immediately made me think of my mom ( who has recently been diligently been interacting with all of my posts here on WordPress. I really appreciate it; since, it should theoretically boost my chances of being seen here through the algorithm. After all, more interaction should lead to better views, right?

Recommender algorithms (better called systems) are complicated. Let’s take a quick overview of them.

Recommender Systems

These things are integral parts of our lives now, and they have been for decades already. You might already know it as “the algorithm” and it decides what you see on YouTube or Pinterest or even your JetPack reader. In order for them to be effective, it’s common sense to make sure your algorithm ensures that the user stays using your product for as long as possible- in other words, you better retain their interest!

One way to do this is to just ask the user. Netflix is a good example.

Did you like Orange Is the New Black?

Then just say so and “the algorithm” will work it’s magic to find items that are similar to that. Or maybe items similar to the collection of your past history of watching and whatever other (useful) data could possibly be gathered.

Now, alternatively, if Netflix shared information between users then they could implement a different type of recommender system. On YouTube, as soon as some of my favorite creators upload I can leave a comment, a like, and even a dislike! This user interaction is sure to affect how other users like me (maybe those with similar subscriptions or interactions) will be recommended that particular video!

Fascinating, right? Of course, it would get kind of expensive to handle that much data for such a complicated system, but I imagine that Google makes enough money from YouTube to handle that part.

Interested in reading more about these kinds of systems? I suggest you check out Manning Publications Practical Recommender Systems.


One response to “Recommender Systems”

  1. I love you and will always be your biggest fan! I love that you put the link to the book in the article as well! Well done!

Leave a commentCancel reply

Exit mobile version