0 views collected around this technical thread.
The award‑winning WSDM 2025 paper reveals how recommendation models magnify popularity bias through spectral alignment of the rating matrix’s leading singular vector with item popularity, and proposes the ReSN regularization technique to mitigate this bias efficiently.
The article examines the rapid rise of AI interview software, highlighting its appeal to job seekers, the potential for cheating, expert criticism, associated risks, and the need for balanced, ethical approaches to ensure fairness in recruitment.
This article examines the challenges of high‑concurrency flash‑sale scenarios in multi‑active architectures, analyzes fairness issues caused by geographic latency, and presents practical solutions such as data‑sharding and global‑clock ordered queues, complemented by a Redis‑based implementation example.
Kuaishou and Tsinghua University were honored with the first‑place Qian Weichang Chinese Information Processing Science and Technology Award for their collaborative content recommendation project, which achieved international‑level innovations in explainable recommendation, bias correction, and edge intelligence, and has been applied widely in Kuaishou's platform and top academic conferences.
The article examines the drawbacks of result‑oriented performance appraisal, arguing that over‑emphasis on outcomes neglects process, fairness, and employee motivation, and proposes balanced evaluation, fair metrics, strategic clarity, organizational efficiency, and performance dashboards to improve management practices.
This article examines the fairness of recommendation systems from both consumer and provider viewpoints, discussing sources of bias, definitions of equality and equity, measurement metrics such as CGF and MMF, and proposes causal embedding models to mitigate unfairness while ensuring sustainable system performance.
This article reviews a series of recent research papers that explore how generative AI, especially large language models, can be integrated into recommender systems to improve next‑basket recommendation, instruction‑following, zero‑shot ranking, fairness evaluation, and generative retrieval paradigms.
PerBid introduces a personalized automated bidding framework that creates context‑aware RL agents for advertiser clusters using a profiling network to embed static and dynamic campaign features, and experiments on Alibaba’s display‑ad platform show up to 10.85% performance gains while markedly improving fairness across heterogeneous advertisers.
This article presents five peer‑reviewed papers accepted to the Kuaishou Community Science track at the 2023 ACM SIGIR conference, covering offline reinforcement learning for recommendation fairness, global residual value for fair item exposure, personalized QoE in short‑form video, multi‑behavior self‑supervised recommendation, and a search‑enhanced sequential recommendation framework.
This article examines the fairness of recommendation systems from both consumer and provider viewpoints, discussing sources of bias, definitions of equality and equity, measurement metrics such as CGF and MMF, causal embedding techniques, experimental results on MovieLens and Yelp, and future research directions.
This article presents a comprehensive overview of trustworthy intelligent decision-making, introducing a decision framework and discussing counterfactual reasoning, complex reward modeling, predictive fairness, and regulatory constraints, while highlighting practical methods and recent research advances in each sub‑area.
The article outlines Ant Group's comprehensive approach to promoting trustworthy AI in large‑scale industrial settings, detailing the four core pillars of robustness, explainability, privacy protection, and fairness, and describing practical methodologies, open platforms, and ecosystem collaborations that drive responsible AI deployment.
This article introduces Java's Semaphore as a thread‑synchronization tool, explains its flow‑control use cases such as database connections and parking lots, provides a complete parking‑lot simulation code sample, and analyzes the underlying AQS‑based implementation including fairness, acquire/release mechanics, and auxiliary methods.
The article details the technical goals, fairness mechanisms, high‑concurrency handling, dynamic routing, peak‑shaving, downstream protection, and on‑site screen interaction strategies used to deliver a seamless multi‑platform live‑stream experience for the Double‑11 Cat Night, while summarizing key takeaways for future large‑scale events.
Alibaba’s 2019 Double‑11 “Cat Evening” live show combined a unified codebase across Youku, Taobao and Tmall with synchronized clocks, latency‑measurement devices and SEI‑injected messages to guarantee fair, zero‑loss interactions, while employing dynamic routing, pre‑warming, peak‑shaving, downstream protection and rehearsed contingency plans to handle massive concurrency and ensure stable, high‑quality user experience.