Seongeun Ryu, Jaesung Hwang, Masoud Reyhani Hamedani, Sang‐Wook Kim · ACM International Conference on Information and Knowledge Management (CIKM) 2026 · 2026
DOI: 10.1145/3799682.3840975
Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).
To alleviate the information overload problem on online news platforms, personalized news recommendation aims to provide news articles aligned with users' interests. Most conventional methods suffer from popularity bias. To mitigate this, popularity-aware methods regard popularity-driven clicks as bias. Although these methods show higher accuracy, they still face three challenges: (C1) discriminating click types on popular news, (C2) topic-unaware popularity modeling, and (C3) dealing with interest signal attenuation. To jointly address these challenges, we propose POPCORN, a novel model-agnostic framework for POPularity-deCOupled inteRest matching in personalized News recommendation, with three key strategies: (1) Popularity-decoupled news modeling to discriminate interest-driven and popularity-driven clicks, (2) Candidate-guided user modeling to amplify a user's interest signal for a candidate news based on its topic, and (3) Popularity-penalized interest matching to suppress popular news not genuinely interesting to the user. Extensive experiments with three real-world datasets and eight state-of-the-art methods demonstrate that the methods augmented with POPCORN consistently outperform their original versions.
ACM International Conference on Information and Knowledge Management (CIKM) 2026 · 0 citations
No comments yet — start the discussion below.