0 views collected around this technical thread.
The paper introduces QIN, a query‑driven user interest network that combines a Relevance Search Unit and a Fused Attention Unit to effectively leverage full‑history user behavior for personalized search, demonstrating significant performance gains in offline benchmarks and online A/B tests.
This article presents JD's end‑to‑end improvements for advertising click‑through‑rate prediction, addressing cold‑start, deep user‑interest mining, and full‑domain collaborative information through a variational feature learning framework, enhanced interest networks (PPNet+, NeNet, Weighted‑MMoE) and exposure‑sequence modeling, achieving over 1% cumulative AUC gain and publication in top conferences.
The article outlines Baidu Haokan’s short‑video recommendation architecture, describing how a unified ranking pipeline uses user‑interest signals, multi‑objective MMOE and deep‑fusion models, and long‑term value estimation to balance personalized user experience, creator exposure, and advertiser goals across billions of daily video plays.
This article explains how a two‑stage recommendation pipeline uses explicit tag‑based recall, inverted indexes, and a multi‑task learning model to improve click‑through and dwell time by dynamically balancing loss weights across tasks.
The SURGE framework leverages graph neural networks to construct and pool interest graphs from user interaction sequences, achieving stable and fast convergence, robust long‑sequence modeling, and significant performance gains over existing sequential recommendation methods on e‑commerce and short‑video datasets.
This article presents the challenges of large‑scale home‑page recommendation at 58.com, describes how behavior‑sequence models such as DIN, DIEN and Transformer are applied and evolved into double‑channel and multi‑channel deep interest architectures, and details offline and online performance optimizations that yielded significant gains in click‑through and conversion rates.