Bird Identification Accuracy in eBird Citizen Science Data (Argentina)
Supplementary materials to quantify the accuracy of photographic bird identifications in eBird Argentina, with per-species ranking and network analysis.
- Size
- 2.7 MB
- Format
- HTML, RAR, XLSX
- License
- CC-BY-4.0
- Access
- Open Access
- Source
- Zenodo (record 6828335)
Your edits to the notebook only affect your own copy in the Hub. Changed it and want the original back? Restore it here.
Challenge
In eBird Argentina's photo identifications, is it the rare species (few vetted photos) or the look-alikes (confusable relatives) that drive misidentification errors, and which species are the worst confusion hubs?
- Where to start
- Load Appendix_S2.xlsx (sheet 'species_rank') and rename 'min (precision, recall)' to accuracy. Rank species by accuracy and split by the 'hard_to_id' flag; plot accuracy against 'samples' on a log scale to test the rarity hypothesis. Then parse the embedded nodes/edges JSON from Appendix_S3.html (the visNetwork widget) to sum confusion weight per species and check how much of it stays within the same 'family'.
- What to share
- A ranked chart of the hardest-to-identify species, an accuracy-vs-sample-size plot (or boxplot by hard_to_id), and a short conclusion on whether rarity or look-alikes explain the errors, backed by the within-family confusion share and the top confusion hubs/pairs.
💬 Discuss this dataset, ask questions and share your results in its discussion thread. Challenges are open-ended — there's no single right answer.