4. Question 4 In recommendation systems, one uses Beta-Gamma threshold learning for trade-off between exploration and exploitation: \theta = \alpha * \theta_{zero} + (1- \alpha) * \theta_{optimal}θ=α∗θzero+(1−α)∗θoptimal. Which of the…
9. Question 9 In Spotify, if a user has indicated himself/herself as youth, then Spotify recommends songs that are most listened by users under 20 years old. What is this…
7. Question 7 Recommendation is one type of Pull mode of information access. 1 point False True
10. Question 10 When adding social network information into recommendation systems, such as friends’ info and friends’ liked items, this can be used to help: 1 point Collaborative filtering…
8. Question 8 In Netflix, if a user has watched a lot of thriller movies, then it recommends “Inception” and “The Silence of the Lambs” to the user. What is…
6. Question 6 Which of the following is not an advantage of Learning to Rank? 1 point Combining multiple sources of features Directly optimize retrieval measures Is…
5. Question 5 Content-based filtering and collaborative filtering can be combined in a recommender system. 1 point False True
3. Question 3 In content-based filtering, an item is recommended to a user based on whether other “similar” users like the item or not. 1 point False True
2. Question 2 Information filtering systems are more suitable to help users satisfy long-term information needs than short-term ad hoc information needs. 1 point True False
1. Question 1 When using the learning to rank framework for combining multiple features into a ranking function, training data composed of queries and relevance judgments is needed to learn…
