Transformer-driven automated analysis of social media narrative structure: An exploration based on sentiment framing and thematic agenda
Rongkang Pei, Zeyu Lyu, Guolong Wang · 2026
Read Transformer-driven automated analysis of social media narrative structure: An exploration based on sentiment framing and thematic agenda by Rongkang Pei, Zeyu Lyu, Guolong Wang (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-0352688.txt.
Description
Rongkang Pei, Zeyu Lyu, and Guolong Wang build an artificial intelligence model for reading social media storms. Their 2026 study tackles narrative structure in posts about public events. Such posts carry opinion, emotion, and fast-moving themes. Older tools treated sentiment, topic, and time separately. The authors fuse all three inside a transformer model built on a RoBERTa language backbone.
The combined model beat four kinds of baseline systems. It recognised narrative structure more accurately. It also classified sentiment and topics better. Performance held up across different events and in noisy, messy text. Those are exactly the conditions of real public opinion monitoring. The authors close with ethics: privacy, anonymisation, and bias all need care. The tool could help officials follow public reaction during crises, provided those safeguards hold.
Catalog record
| CREATOR | Rongkang Pei, Zeyu Lyu, Guolong Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78260 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |