Skip to main content
[Preprint]. 2022 Mar 21:2021.02.09.21251454. Originally published 2021 Feb 12. [Version 2] doi: 10.1101/2021.02.09.21251454

Table 4:

Comparison of single and multi-corpus training on NER datasets. SOTA performance on the CADEC dataset could not be determined as other works use different entity subsets as symptoms. SOTA performance on Micromed was reported on the full dataset, however, as of this work, only 51% of the tweets are available.

Training Set SOTA
P/R/F1 		Single-corpus
P/R/F1 		Multi-corpus
P/R/F1
DS-NER 0.86/0.78/0.82 0.88/0.82/0.85 0.88/0.84/0.86
Tw-NER 0.76/0.68/0.72 0.78/0.70/0.74 0.79/0.71/0.75
CADEC - 0.75/0.81/0.78 0.74/0.80/0.77
Micromed 0.79/0.66/0.72* 0.65/0.62/0.64 0.63/0.60/0.62
TwiMed - 0.57/0.68/0.62 0.58/0.68/0.63