A pretrained biomedical large language model for Persian biomedical text mining
Baqer M. Merzah, Tania Taami, Salman Asoudeh, Saeed Mirzaee, Amir Reza Hossein Pour, Amir Ali Bengari, Ehsan Karimi Shahmarvandi · 2026
Read A pretrained biomedical large language model for Persian biomedical text mining by Baqer M. Merzah, Tania Taami, Salman Asoudeh, Saeed Mirzaee, Amir Reza Hossein Pour, Amir Ali Bengari, Ehsan Karimi Shahmarvandi (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/PMC13518823.txt.
Description
Baqer M. Merzah leads a seven-author team introducing BioPars in 2026, a large language model built for Persian biomedical text. General chatbots remember medical facts well yet stumble on deeper real-world questions, and Persian medical question answering had no dedicated model. The team assembled training material from over 10,000 articles, textbooks, and medical websites.
Their benchmark, BioParsQA, holds 5,231 Persian medical questions and answers. BioPars scored a ROUGE-L of 29.99 on that set, edging GPT-4 by 1.0 point in their reporting, with a BERTScore of 90.87 using one ranking method. MoverScore and BLEURT also topped the three rival models tested. Fine-grained reasoning, the authors note, still needs work.
Catalog record
| CREATOR | Baqer M. Merzah, Tania Taami, Salman Asoudeh, Saeed Mirzaee, Amir Reza Hossein Pour, Amir Ali Bengari, Ehsan Karimi Shahmarvandi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 94669 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |