Over the last several years, there has been a trend in recommendation systems shifting from COLLABORATIVE FILTERING to CONTENT-BASED FILTERING.
Each workflow are the following.
Let’s say we want to recommend a restaurant to a user.
Collaborative Filtering
1. looks at User
2. Find Similar User
3. Recommend what similar people liked
Content-Based Filtering
1. Look at User
2. Look at Menu in a restaurant
3. Look other information about restaurant
4. Decide if restaurant is a match
What’s great about Content-based filtering is that even if a menu is something new and nobody hasn’t yet tried it out, it would still understand how well of a match it is to you by considering other elements.
