Understanding content-based recommender systems

One of the most common and successful types of recommendations are content-based. The core idea is that if I expressed a preference for a certain set of items, I will most likely be interested in more items that share the same attributes. For example, the fact that I watched Finding Nemo (2003) can be used as an indication that I will be interested in other movies from the animation and comedy genres.

Alternatively, watching one of the original Star Wars movie can be interpreted as a signal that I like other movies from the franchise, or movies with Harrison Ford, or directed by George Lucas, or science fiction in general. Indeed, Netflix employs such an algorithm, except at a more granular ...

Get Julia Programming Projects now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.