GateSID: Adaptive Gating for Semantic-Collaborative Alignment in Cold-Start Recommendation DOI: XXXXXXX.XXXXXXX Conference: GateSID; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Information systems Recommender systems

Hai Zhu, Yantao Yu, Lei Shen, Bing Wang*, Xiaoyi Zeng · 2026

Read GateSID: Adaptive Gating for Semantic-Collaborative Alignment in Cold-Start Recommendation DOI: XXXXXXX.XXXXXXX Conference: GateSID; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Information systems Recommender systems by Hai Zhu, Yantao Yu, Lei Shen, Bing Wang*, Xiaoyi Zeng (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-22916.txt.

Description

Zhu and co-authors present GateSID, a recommendation method for cold-start items. New items lack the user signals that power standard ranking. Adding semantic data helps, but leaning on it too hard can blur real user patterns. GateSID uses an adaptive gate to weigh semantic and collaborative signals by item maturity. Features are first coded into hierarchical semantic IDs. Two parts then fuse attention and align signals, with stronger alignment for cold items. Industry tests and an online trial show gains in sales and clicks.

Catalog record

CREATORHai Zhu, Yantao Yu, Lei Shen, Bing Wang*, Xiaoyi Zeng
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18414
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22916

Browse the full library · About the Tower of Babel