Item-Based Collaborative FilteringΒΆ
Python notebook: https://github.com/daviskregers/data-science-recap/blob/main/20-item-based-movie-similarity.ipynb
- What if we based recommendations on relationships between things instead of people?
- A movie will always be the same movie - it doesn't change.
- There are usually fewer things than people (less computation to do)
- Harder to game the system
- Find every pair of movies that were watched by the same person
- Measure the similarity of their ratings accross all users who watched both
- Sort by movie, then by similarity strength
(this is just one way to do it.)