Proceedings of the 2007 ACM Conference on Recommender Systems 2007
DOI: 10.1145/1297231.1297251
|View full text |Cite
|
Sign up to set email alerts
|

Improving new user recommendations with rule-based induction on cold user data

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
30
0
1

Year Published

2009
2009
2022
2022

Publication Types

Select...
3
3
2

Relationship

0
8

Authors

Journals

citations
Cited by 45 publications
(31 citation statements)
references
References 10 publications
0
30
0
1
Order By: Relevance
“…Including ratings before this time aims to remove the system-wide cold-start problem that recommender systems face, i.e. a state where predictions can not be made as there are little to no historical ratings available [7]. In this work, we set = 500 days from the first rating in the dataset, and our final subsets have about 60, 000 users: setting the value as we did is equivalent to bootstrapping a recommender system with 10, 000 users.…”
Section: Temporal Experimentsmentioning
confidence: 99%
“…Including ratings before this time aims to remove the system-wide cold-start problem that recommender systems face, i.e. a state where predictions can not be made as there are little to no historical ratings available [7]. In this work, we set = 500 days from the first rating in the dataset, and our final subsets have about 60, 000 users: setting the value as we did is equivalent to bootstrapping a recommender system with 10, 000 users.…”
Section: Temporal Experimentsmentioning
confidence: 99%
“…Moreover, rule-based induction [5] on cold user data has been also used to improve new user recommendation. It combines with various-community spaces which relate to cold user data (age, occupation, location, etc.).…”
Section: Related Workmentioning
confidence: 99%
“…Then, the set B' can be obtained combined with the cliques and the set B'' further can be formed according to preferential attachment characteristic. Finally, the neighbor set of new user is formed according to the equation (5).…”
Section: Datasetmentioning
confidence: 99%
See 1 more Smart Citation
“…This paper also focuses on binary Web usage data, however it differentiates between different types of usage information. Therefore, the approach is also related to the α-community spaces model, where different user features termed α i are employed for determining similar users [15]. Nguyen et al determine clusters of similar users according to feature α i and use a rule-based induction approach to derive recommendations from the corresponding cluster.…”
Section: Introductionmentioning
confidence: 99%