Essentially, this is about creating alternative training data to "seed" a recommender engine - if the engine has no previous data about a video series, what oth

Essentially, this is about creating alternative training data to "seed" a recommender engine - if the engine has no previous data about a video series, what other data (cover art) can be used to recommend a given viewer to watch or not watch? Is a recommender a type of Machine Learning?