From Events to Trending: A Multi-Stage Hotspots Detection Method Based on Generative Query Indexing

Kaichun Wang * , Yanguang Chen * , Ting Zhang, Mengyao Bao, Keyu Chen, Xu Hu, and Yongliang Wang, Jingsheng Yang, Jinsong Zhang, Fei Lu · 2026

Read From Events to Trending: A Multi-Stage Hotspots Detection Method Based on Generative Query Indexing by Kaichun Wang * , Yanguang Chen * , Ting Zhang, Mengyao Bao, Keyu Chen, Xu Hu, and Yongliang Wang, Jingsheng Yang, Jinsong Zhang, Fei Lu (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-2601-05258.txt.

Description

Kaichun Wang and 9 co-authors detect trending topics inside conversational AI systems in a 2026 paper. Chatbots have become a popular gateway to information, yet most handle news-related trending queries poorly, and user experience suffers when a breaking topic is treated like an ordinary question. Trending detection built for traditional search engines underperforms here, because conversational queries differ radically in distribution and expression from typed search queries.

Their multi-stage method, based on generative query indexing, runs from raw events to trending hotspots suited to dialogue traffic. Detecting a surge early lets a conversational system route it to fresher sources or special handling. The work targets a gap the authors note is largely unexplored for dialogue systems specifically.

Catalog record

CREATORKaichun Wang * , Yanguang Chen * , Ting Zhang, Mengyao Bao, Keyu Chen, Xu Hu, and Yongliang Wang, Jingsheng Yang, Jinsong Zhang, Fei Lu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-05258

Browse the full library · About the Tower of Babel