A Simple Convolutional Generative Network for Next Item Recommendation
- Submitting institution
-
University of Glasgow
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11-12080
- Type
- E - Conference contribution
- DOI
-
10.1145/3289600.3290975
- Title of conference / published proceedings
- Twelfth ACM International Conference on Web Search and Data Mining
- First page
- 582
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- -
- Year of publication
- 2019
- URL
-
http://eprints.gla.ac.uk/182377/
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
4
- Research group(s)
-
-
- Citation count
- 30
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- ORIGINALITY: Developed an effective generative model for learning high-level representation from both short and long-range item dependencies, a major issue (we demonstrate problems with long-range dependencies in prior approaches) in session-based recommendation. SIGNIFICANCE: Developed in collaboration with Telefonica, and incorporated into their product suites and later by Tencent. Published in one of the top conferences in information retrieval, with an acceptance rate of 16.4% and very highly cited already. RIGOUR: The techniques presented in this paper are implemented and carefully tested on two benchmarks, obtaining world-leading performance (RecSys 2015 challenge data set, Last.fm).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -