BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:AI in Metascience
DTSTART:20260617T070000Z
DTEND:20260619T160000Z
DTSTAMP:20260606T130400Z
UID:indico-event-3975@indico.uni-muenster.de
DESCRIPTION:Speakers: Lukas Röseler\, Lukas Wallrich (Birkbeck\, Universi
 ty of London)\, Tom Heyman\, Yingjie Li\, Isis Trajano\n\nA focused online
  event for applying AI to metascience—the study and improvement of how s
 cience is done.\nAims\n\nClarify AI vs. LLMs and where each fits in metasc
 ience.\nEnable ethical\, transparent\, and effective AI use.\nProvide prac
 tical methods for coding\, data extraction\, and analysis.\nReduce duplica
 tion\, improve coordination\, and build capacity for funding calls.\n\n \
 nWho should attend\n\nMetascience researchers\, methodologists\, meta‑re
 search students/postdocs\, research software engineers\, librarians/inform
 ation specialists\, editors\, and research integrity/ethics leads.\n\n \n
 Scope\n\nPlenary: AI capabilities\, limits\, and evaluation for metascienc
 e.\nWorkflow: From literature to dataset—deduplication\, entity resoluti
 on\, provenance.\nLLMs in practice: Screening and data extraction with aud
 it trails.\nBeyond LLMs: Retrieval\, weak supervision\, and classical ML f
 or study classification.\nReproducible pipelines: Containers\, workflow or
 chestration\, and transparency by design.\n\n\nhttps://indico.uni-muenster
 .de/event/3975/
LOCATION:Online
URL:https://indico.uni-muenster.de/event/3975/
END:VEVENT
END:VCALENDAR