When is content-based more appropriate than collaborative filtering?

Question

I know the difference between content-based and collaborative filtering approach in recommender systems. I also know some of the articles said collaborative filtering have some advantages than content-based, some of them also suggest to use both method (hybrid) to make a better system recommendation.

Is there a specific case where the use of one method (content-based, specifically) is better than another? Because if there is no case at all, why both methods are considered to be on the same "level", why not focus on just one method? For example, focus on collaborative filtering or hybrid method (as an extension for collaborative filtering).

Answer 1

Some of the cases content-based filtering is useful is:

• Cold-start problem: it happens when no previous information about user history is available to build collaborative filtering, so in this case, we offer to the user some items then recommend based on the similarity between these items and other items in the dataset alternate of recommending any items that maybe not with the user taste.

• Transparency: collaborative method gives you the recommendation because some unknown users have the same taste as you that cause a problem if your data is biased towards one taste makes new user haven't enough similar users have the same taste, but the content-based method can tell you they recommend the items based on what features, that helps you to determine which factors affect in the recommendation.