I've very new to data science and am not sure what terms to even search for, so please forgive any ignorance.
Let's assume that I'm building a movie recommendation engine. A user will rate a subset of movies and this information will be used to recommend new movies for them to watch. I can have a person either rate each movie they've seen on a scale of 1-10 (or 1-5). Or I can have them simply choose "like" or "dislike."
If I recall correctly, Netflix started with the numeric scale and then migrated to the binary one. What are the advantages and disadvantages to both?