Table 11 Examples of errors and predictions of each neural model (B: BERT; F: Flair; SL: SequenceLabeler)

Ambiguity

el nivel de fósforo se redujo (‘phosphorus level decreased’)

Prediction:

O B-PROC I-PROC I-PROC O O

✗: B, F, SL

Reference:

O O O B-CHEM O O

grupos de 20 miembros (‘20-member groups’)

Prediction:

O O O B-ANAT

✗: SL

Reference:

O O O O

✓: B, F

FNs

Lurbinectedin ( PM01183 ) (‘Lurbinectedin (PM01183)’)

Prediction:

O O O O

✗: SL

Reference:

B-CHEM O B-CHEM O

✓: B, F

episodios de NF (‘episodes of FN’ [‘febrile neutropenia’])

Prediction:

O O O

✗: B, SL

Reference:

O O B-DISO

✓: F

FPs

gen AVXS-101 (‘AVXS-101 gene’)

Prediction:

O B-CHEM

✗: B, SL

Reference:

O O

✓: F

estudio BREATH-19 (‘BREATH-19 study’)

Prediction:

O B-DISO

✗: SL

Reference:

O O

✓: B, F

Scope

eventos adversos graves (‘severe adverse events’)

Prediction:

B-DISO I-DISO I-DISO

✗: B, SL

Reference:

B-DISO I-DISO O

✓: F

cirugía ginecológica abierta (‘open gynecologic surgery’)

Prediction:

B-PROC I-PROC I-PROC

✗: B, F, SL

Reference:

B-PROC I-PROC O

estudios comparativos de la eficacia (‘compared efficacy studies’)

Prediction:

B-PROC O O O O

✗: SL

Reference:

B-PROC I-PROC I-PROC I-PROC I-PROC

✓: B, F