The basic intuition of IBCF is that, given a new user, the algorithm considers the user's purchases and recommends similar items through the following steps:
- For two given items, measure how similar they are in terms of having received similar ratings by similar users.
- Next, find the k-most similar items for each given item.
- Finally, for each user, identify the items that are most similar to the user's purchases.
The basic idea of UBCF is that, given a new user, we will identify its similar users and then recommend the top-rated items purchased by similar users through the following steps:
- Measure how similar each user is to the new one using popular similarity measures such as correlation and cosine.
- Next, identify the ...