Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.20126814 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866902199211130880 |
|---|---|
| author | Reyes-Consuelo, Alejandro Kiss, Jocelyne Voisin, Julien |
| author_facet | Reyes-Consuelo, Alejandro Kiss, Jocelyne Voisin, Julien |
| contents | <p>metaScreener is an open-source, cross-platform desktop application implemented in<br>Python/Tkinter that delivers a sequential, multi-stage pipeline for systematic literature<br>screening. The software integrates deterministic rule-based filters with large<br>language model (LLM) inference to automate high-volume citation screening within<br>structured evidence synthesis workflows. Its plugin architecture exposes seven interoperable<br>modules covering citation extraction, reference resolution, eligibility<br>criteria structuring, heuristic-based filtering, and LLM-based full-record eligibility<br>adjudication. Each screening decision is logged within a timestamped, SHA-256-<br>verified bundle archive that satisfies the audit and reproducibility requirements<br>expected in rigorous evidence synthesis methodology. In a demonstration use case<br>comprising 776 candidate records, the pipeline reduced the corpus to 73 records requiring<br>full human review — a 90.6% reduction — with deterministic pre-filtering<br>accounting for 98.3% of exclusions and LLM-assisted stages providing fine-grained<br>adjudication over the residual candidate set. Complete decision traceability was<br>preserved across all stages. metaScreener is released under the MIT licence and is<br>publicly available at <a href="https://github.com/lars-ulaval/metaScreener">https://github.com/lars-ulaval/metaScreener.</a></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_20126814 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | metaScreener: A Plugin-Based Desktop Application for Human-in-the-Loop Systematic Literature Screening Reyes-Consuelo, Alejandro Kiss, Jocelyne Voisin, Julien <p>metaScreener is an open-source, cross-platform desktop application implemented in<br>Python/Tkinter that delivers a sequential, multi-stage pipeline for systematic literature<br>screening. The software integrates deterministic rule-based filters with large<br>language model (LLM) inference to automate high-volume citation screening within<br>structured evidence synthesis workflows. Its plugin architecture exposes seven interoperable<br>modules covering citation extraction, reference resolution, eligibility<br>criteria structuring, heuristic-based filtering, and LLM-based full-record eligibility<br>adjudication. Each screening decision is logged within a timestamped, SHA-256-<br>verified bundle archive that satisfies the audit and reproducibility requirements<br>expected in rigorous evidence synthesis methodology. In a demonstration use case<br>comprising 776 candidate records, the pipeline reduced the corpus to 73 records requiring<br>full human review — a 90.6% reduction — with deterministic pre-filtering<br>accounting for 98.3% of exclusions and LLM-assisted stages providing fine-grained<br>adjudication over the residual candidate set. Complete decision traceability was<br>preserved across all stages. metaScreener is released under the MIT licence and is<br>publicly available at <a href="https://github.com/lars-ulaval/metaScreener">https://github.com/lars-ulaval/metaScreener.</a></p> |
| title | metaScreener: A Plugin-Based Desktop Application for Human-in-the-Loop Systematic Literature Screening |
| url | https://doi.org/10.5281/zenodo.20126814 |