Podcast
Questions and Answers
In collaborative filtering, what is the goal of recommending items to customers?
In collaborative filtering, what is the goal of recommending items to customers?
- To recommend items randomly to increase diversity
- To recommend items that have previously been purchased by similar customers (correct)
- To recommend items based on their popularity
- To recommend items based on users who are most different
What similarity measure does Amazon use in generating recommendations?
What similarity measure does Amazon use in generating recommendations?
- Pearson correlation
- Euclidean distance
- Cosine similarity (correct)
- Jaccard index
What kind of data does Amazon use to generate recommendations?
What kind of data does Amazon use to generate recommendations?
- Only product features
- Only user demographic data
- Both user demographic data and product features
- Only user interaction data without looking at features (correct)
Which users are considered more similar in collaborative filtering?
Which users are considered more similar in collaborative filtering?
What is the binary representation of items purchased by user u?
What is the binary representation of items purchased by user u?
In the context of defining similarity between users and items, what does 𝑅.,+ represent?
In the context of defining similarity between users and items, what does 𝑅.,+ represent?
How is the best case scenario defined for similarity scores in terms of Euclidean distance?
How is the best case scenario defined for similarity scores in terms of Euclidean distance?
What defines the Euclidean distance between two items i and j in this context?
What defines the Euclidean distance between two items i and j in this context?
What is the purpose of subtracting the average rating by user v from each rating in the formula for Pearson correlation similarity between users?
What is the purpose of subtracting the average rating by user v from each rating in the formula for Pearson correlation similarity between users?
How does the cosine similarity differ from Pearson correlation similarity in terms of item consideration?
How does the cosine similarity differ from Pearson correlation similarity in terms of item consideration?
Why is it important to consider only shared items when determining similarity between users based on ratings?
Why is it important to consider only shared items when determining similarity between users based on ratings?
In the context of recommendation systems, what is the purpose of Euclidean distance in relation to similarities between items?
In the context of recommendation systems, what is the purpose of Euclidean distance in relation to similarities between items?
Flashcards are hidden until you start studying
