A multi-task learning-based fully connected neural network for personalized news recommendation
ZhuoMin Ren, Hong Xie · 2026
Read A multi-task learning-based fully connected neural network for personalized news recommendation by ZhuoMin Ren, Hong Xie (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/PMC13542204.txt.
Description
ZhuoMin Ren and Hong Xie propose, in a 2026 study, a neural network that recommends news by learning two jobs at once: predicting clicks and learning each reader's spread of interests. Standard recommenders model shifting tastes poorly, blur accuracy against variety, and stumble with brand-new users.
Their model embeds each user's behaviour sequence so taste drift becomes visible, then trains both tasks in one shared representation. On the public Microsoft News Dataset it beat collaborative filtering, content-based methods, a recurrent rival, and transformer models, improving ranking and click measures by roughly nine to twelve percent. Cold-start accuracy for new users also rose by about eleven percent, with repeated runs confirming stability.
Catalog record
| CREATOR | ZhuoMin Ren, Hong Xie |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 135348 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |