Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

MetaBeeAI : An AI pipeline for structured evidence extraction from biological literature

Parkinson, Rachel H. ; Cerbone, Henry ; Mieskolainen, Mikael ; Cao, Shuxiang ; Wilson, Alasdair D. ; Albacete, Sergio ; Armstrong, Emily B. ; Bass, Chris ; Botías, Cristina and Brown, Andrew , et al. (2026) In Ecological Informatics 96.
Abstract

The volume and complexity of scientific literature are expanding rapidly, making it increasingly difficult to extract and synthesize information across studies. This challenge is particularly acute in the biological sciences, where evidence spans multiple levels of organization and heterogeneous experimental designs. Large Language Model (LLM) pipelines offer a scalable route to evidence synthesis, but many existing approaches lack transparency, modularity, and effective mechanisms for human oversight. We present MetaBeeAI, an open-source, modular pipeline that integrates established LLM techniques into a coherent, auditable workflow for structured data extraction in biology. MetaBeeAI combines modular prompting, multi-pass extraction,... (More)

The volume and complexity of scientific literature are expanding rapidly, making it increasingly difficult to extract and synthesize information across studies. This challenge is particularly acute in the biological sciences, where evidence spans multiple levels of organization and heterogeneous experimental designs. Large Language Model (LLM) pipelines offer a scalable route to evidence synthesis, but many existing approaches lack transparency, modularity, and effective mechanisms for human oversight. We present MetaBeeAI, an open-source, modular pipeline that integrates established LLM techniques into a coherent, auditable workflow for structured data extraction in biology. MetaBeeAI combines modular prompting, multi-pass extraction, and expert-in-the-loop validation within an interface that presents model outputs alongside source text, enabling inspection, correction, and iterative refinement. The pipeline produces machine-readable records of prompts, configurations, and expert annotations, supporting reproducibility and continuous improvement. We apply MetaBeeAI to 924 research papers on bees and pesticides, extracting structured information on species, compounds, exposure designs, and experimental context. Evaluation demonstrates improved consistency, convergence with expert judgement, and robustness across heterogeneous biological studies, highlighting the value of expert-guided refinement. MetaBeeAI provides a transparent and extensible framework for scalable evidence synthesis, supporting reliable integration of LLMs into biological research workflows.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; and (Less)
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Expert-in-the-loop, Interpretable AI/ML, Knowledge extraction, Large Language Models (LLM), Systematic review
in
Ecological Informatics
volume
96
article number
103813
publisher
Elsevier
external identifiers
  • scopus:105039801046
ISSN
1574-9541
DOI
10.1016/j.ecoinf.2026.103813
language
English
LU publication?
yes
id
ac11b294-cf07-477d-a2e9-83fc02ef9087
date added to LUP
2026-07-02 14:11:03
date last changed
2026-07-02 17:43:45
@article{ac11b294-cf07-477d-a2e9-83fc02ef9087,
  abstract     = {{<p>The volume and complexity of scientific literature are expanding rapidly, making it increasingly difficult to extract and synthesize information across studies. This challenge is particularly acute in the biological sciences, where evidence spans multiple levels of organization and heterogeneous experimental designs. Large Language Model (LLM) pipelines offer a scalable route to evidence synthesis, but many existing approaches lack transparency, modularity, and effective mechanisms for human oversight. We present MetaBeeAI, an open-source, modular pipeline that integrates established LLM techniques into a coherent, auditable workflow for structured data extraction in biology. MetaBeeAI combines modular prompting, multi-pass extraction, and expert-in-the-loop validation within an interface that presents model outputs alongside source text, enabling inspection, correction, and iterative refinement. The pipeline produces machine-readable records of prompts, configurations, and expert annotations, supporting reproducibility and continuous improvement. We apply MetaBeeAI to 924 research papers on bees and pesticides, extracting structured information on species, compounds, exposure designs, and experimental context. Evaluation demonstrates improved consistency, convergence with expert judgement, and robustness across heterogeneous biological studies, highlighting the value of expert-guided refinement. MetaBeeAI provides a transparent and extensible framework for scalable evidence synthesis, supporting reliable integration of LLMs into biological research workflows.</p>}},
  author       = {{Parkinson, Rachel H. and Cerbone, Henry and Mieskolainen, Mikael and Cao, Shuxiang and Wilson, Alasdair D. and Albacete, Sergio and Armstrong, Emily B. and Bass, Chris and Botías, Cristina and Brown, Andrew and Hayward, Angela J. and Herbertsson, Lina and Jones, Andrew K. and Nagloo, Nicolas and Nicholls, Elizabeth and Rigosi, Elisa and Sgolastra, Fabio and Siviter, Harry and Stanley, Dara A. and Straub, Lars and Straw, Edward A. and Tadei, Rafaela and Walter, Kieran and Stevance, Heloise F. and Daniels, Ryan K. and Lambert, Ben and Roberts, Stephen}},
  issn         = {{1574-9541}},
  keywords     = {{Expert-in-the-loop; Interpretable AI/ML; Knowledge extraction; Large Language Models (LLM); Systematic review}},
  language     = {{eng}},
  publisher    = {{Elsevier}},
  series       = {{Ecological Informatics}},
  title        = {{MetaBeeAI : An AI pipeline for structured evidence extraction from biological literature}},
  url          = {{http://dx.doi.org/10.1016/j.ecoinf.2026.103813}},
  doi          = {{10.1016/j.ecoinf.2026.103813}},
  volume       = {{96}},
  year         = {{2026}},
}