Phenonizer: A Fine-Grained Phenotypic Named Entity Recognizer for Chinese Clinical Texts

Table 2

The number of various entities in benchmark datasets.

Entity	TCM-HN	COVID-19	TCM-HB
Presented symptom	753,541	60,364	170,047
Negated symptom	469,142	31,092	126,359
Disease	55,857	—	55,783
Tongue and pulse	2,621	1,497	—
Body parts	1,042	—	—
Operation	959	—	8,540
Date	—	18,653	—
Duration of symptoms	—	9,042	23,276
Past history	—	5,796	—
Inducement	—	3,524	—
Drug	—	3,376	23,191
Frequency	—	—	16,127
Principle	—	—	11,993