Forecasting user engagement and competing cascades in social media diffusion: A Hawkes-Transformer approach
Wen Zhang, Zhe Jing, Yue Guo · 2026
Read Forecasting user engagement and competing cascades in social media diffusion: A Hawkes-Transformer approach by Wen Zhang, Zhe Jing, Yue Guo (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/plos-pone-0354472.txt.
Description
Zhang and colleagues, in a 2026 study, model how posts spread on social media, using data from Sina Weibo, a major Chinese platform. Likes, comments, and reposts do not just add up; each post competes with others on the same hot topic for limited attention. Forecasting that contest helps explain how online attention moves.
Their model blends two tools. A self-exciting process, which captures how each burst of activity sparks more, estimates how strongly posts excite or drain each other. Those estimates then feed a transformer, the deep learning design behind modern language tools, which makes the forecast. Reposts strongly amplified spread. Comments, by contrast, could slow a post by pulling attention away. Parallel posts on one topic tended to split attention rather than boost each other. The authors say the blend keeps the forecast sharp while staying readable, a help for anyone studying how platforms steer public attention.
Catalog record
| CREATOR | Wen Zhang, Zhe Jing, Yue Guo |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 87999 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |