On the GeoEDdA corpus of 18th century French encyclopedias, fine-tuned CamemBERT achieves the highest macro-averaged F1, closely followed by Flair, and few-shot GPT models lag behind but may help when labeled data are scarce.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Token and Span Classification for Entity Recognition in French Historical Encyclopedias
On the GeoEDdA corpus of 18th century French encyclopedias, fine-tuned CamemBERT achieves the highest macro-averaged F1, closely followed by Flair, and few-shot GPT models lag behind but may help when labeled data are scarce.