Expert identification blitz: A rapid high value approach for assessing and improving iNaturalist identification accuracy and data precision and confidence

Societal Impact Statement Citizen science data are increasingly used in research and conservation, so assessing and improving data accuracy is important. We recruited 50 experts to review a dataset of Western Australian plant records from iNaturalist. Across three weeks, almost 11,000 records received at least one identification. Of the 7,000 records initially identified to a species level, 92% were correctly identified. Building these banks of expert‐curated data and images is crucial for biodiverse regions with many endemic and threatened species, and regions without a comprehensive Flora, as reliably‐identified occurrence records that supplement herbarium vouchers help define species distributions and better inform conservation assessments. Summary Citizen science data now make up a significant portion of global biodiversity records. The utility of these data for research and conservation strongly depends on identification accuracy, which varies across taxa, underscoring the importance of expert engagement. We used a large iNaturalist dataset of vascular plant records from three regions in Southwest Western Australia as a case study, organising a three‐week expert identification event to assess and enhance identification accuracy and data precision and confidence. Before the event, Research Grade records comprised 48% of our species‐level dataset and had an accuracy rate of 97%, while 52% of species‐level records were marked as needs ID, of which 89% were confirmed to have been accurately identified. The event led to a net increase of almost 250% for Research Grade records for our dataset, and a net increase of 183 new species recorded in iNaturalist for the three focal regions. This collaboration between experts and citizen scientists provides a way to quantify identification error rates for downstream statistical analysis and improves uncertain or incorrect identifications. It contributes to a growing collection of confidently identified photographs for the Western Australian flora, which serve as valuable resources for citizen scientists learning about their local flora and collecting further data. This process is a vital component of a virtuous cycle, continually improving data in one of the world's biodiversity hotspots. Our approach is globally applicable to any flora, requiring only the selection of a region and a manageable dataset and the recruitment of relevant regional experts.

Authors

Institutions

Publication Details

Journal
Plants People Planet
Published
2025-03-16
DOI
https://doi.org/10.1002/ppp3.70005
Citations
4
Primary Topic
Big Data and Business Intelligence
Type
article
Field-Weighted Citation Impact
9.14
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Expert identification blitz: A rapid high value approach for assessing and improving iNaturalist identification accuracy and data precision and confidence

William K. Cornwell, Hervé Sauquet, Kelly Anne Shepherd, Russell L. Barrett et al.
4 citations
Plants People Planet
Big Data and Business Intelligence
9.14
article

Expert identification blitz: A rapid high value approach for assessing and improving iNaturalist identification accuracy and data precision and confidence

William K. Cornwell, Hervé Sauquet, Kelly Anne Shepherd, Russell L. Barrett, Juliet Ann Wege, Thomas Mesaglio
article en
4 citations

Abstract

Societal Impact Statement Citizen science data are increasingly used in research and conservation, so assessing and improving data accuracy is important. We recruited 50 experts to review a dataset of Western Australian plant records from iNaturalist. Across three weeks, almost 11,000 records received at least one identification. Of the 7,000 records initially identified to a species level, 92% were correctly identified. Building these banks of expert‐curated data and images is crucial for biodiverse regions with many endemic and threatened species, and regions without a comprehensive Flora, as reliably‐identified occurrence records that supplement herbarium vouchers help define species distributions and better inform conservation assessments. Summary Citizen science data now make up a significant portion of global biodiversity records. The utility of these data for research and conservation strongly depends on identification accuracy, which varies across taxa, underscoring the importance of expert engagement. We used a large iNaturalist dataset of vascular plant records from three regions in Southwest Western Australia as a case study, organising a three‐week expert identification event to assess and enhance identification accuracy and data precision and confidence. Before the event, Research Grade records comprised 48% of our species‐level dataset and had an accuracy rate of 97%, while 52% of species‐level records were marked as needs ID, of which 89% were confirmed to have been accurately identified. The event led to a net increase of almost 250% for Research Grade records for our dataset, and a net increase of 183 new species recorded in iNaturalist for the three focal regions. This collaboration between experts and citizen scientists provides a way to quantify identification error rates for downstream statistical analysis and improves uncertain or incorrect identifications. It contributes to a growing collection of confidently identified photographs for the Western Australian flora, which serve as valuable resources for citizen scientists learning about their local flora and collecting further data. This process is a vital component of a virtuous cycle, continually improving data in one of the world's biodiversity hotspots. Our approach is globally applicable to any flora, requiring only the selection of a region and a manageable dataset and the recruitment of relevant regional experts.

Plants People PlanetVol. 7(5)
UNSW Sydney (AU), Royal Botanic Garden Sydney (AU), Environmental Earth Sciences (AU)
Openalex Percentile: Top 3%
Big Data and Business Intelligence
9.14
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.