← Back to Computation and Language
cs.CL

Can AI automatically turn research papers into evidence summaries?

Massimiliano Pronesti, Angelo Miculescu, Mohsin Kapdi, Paul Flanagan, Oisín Redmond, Joao Bettencourt-Silva, Gurdeep Mannu, Spiros Denaxas, Rui Bebiano Da Providencia E Costa, Anya Belz, Yufang Hou

June 1, 2026

Systematic reviews require researchers to manually extract outcome numbers from dozens of papers, define comparable interventions, and feed structured data into specialized software—a bottleneck that takes weeks. AutoForest automates the entire pipeline from raw PDFs to publication-ready forest plots using language models for data extraction and statistical synthesis. Clinician testing showed it substantially speeds up evidence synthesis and makes meta-analyses accessible without specialized domain training.
Published as AutoForest: Automatically Generating Forest Plots from Biomedical Studies with End-to-End Evidence Extraction and Synthesis arXiv:2606.02403
Read the original paper →