Quote:
Originally Posted by brandonstills
I was thinking it would work great for automatically optimizing content to get the most clicks. Maybe clustering users by their preferences to prevent more relevant content. Not sure how feasible it is though since I only know the very basics of machine learning.
|
There are a lot of ways to use machine learning algorithms but to create one is always a tedious process. That's why it's very expensive.
Netflix apparently spends $150 million a year on their recommendation system.
You can check this article here.
I know the programmers who developed a similar algorithm for LiveJasmin a few years ago. LiveJasmin has deep packets so they can afford to pay for this type of development.
I often see some very simple "recommendation systems" on adult VOD sites and content stores that don't require any machine learning algorithm. Tha's a different story. These type of systems are not expensive. I think most small sites don't really need anything more complex.