. 2022 Jun 29;6(6):e34834. doi: 10.2196/34834

Table 4.

Comparison of the performance of machine learning models for detecting the accuracy of health-related tweets.

Model and class			Precision		Recall		F₁ score		Macroaverage		Model accuracy
AraBERTv2^a-base
	Inaccurate	0.804		0.7627		0.7826		0.8279		0.8397
	Accurate	0.86		0.8866		0.8731		0.8279		0.8397
AraBERTv2-large
	Inaccurate	0.8276		0.8136^b		0.8205^b		0.8564^b		0.8654^b
	Accurate	0.8878		0.8969		0.8923		0.8564^b		0.8654^b
AraBERTv0.2^c-base
	Inaccurate	0.8519		0.7797		0.8142		0.8543^b		0.8654^b
	Accurate	0.8725		0.9175		0.8945		0.8543^b		0.8654^b
AraBERTv0.2-large
	Inaccurate	0.8448		0.8305^d		0.8376^d		0.8701^d		0.8782^d
	Accurate	0.898^d		0.9072		0.9025^d		0.8701^d		0.8782^d
MARBERT
	Inaccurate	0.7759		0.7627		0.7692		0.8154		0.8269
	Accurate	0.8571		0.866		0.8615		0.8154		0.8269
ARBERT
	Inaccurate	0.7903		0.8305^d		0.8099		0.8447		0.8526
	Accurate	0.8936		0.866		0.8796		0.8447		0.8526
QARiB
	Inaccurate	0.7797		0.7797		0.7797		0.8228		0.8333
	Accurate	0.866		0.866		0.866		0.8228		0.8333
ArabicBERT^e-large
	Inaccurate	0.8654		0.7627		0.8108		0.8532		0.8654^b
	Accurate	0.8654		0.9278^b		0.8955^b		0.8532		0.8654^b
ArabicBERT-base
	Inaccurate	0.8913^d		0.6949		0.781		0.83492		0.8525
	Accurate	0.8364		0.9485^d		0.8889		0.83492		0.8525
BLSTM^f Mazajak CBOW^g
	Inaccurate	0.7719		0.7458		0.7586		0.8079		0.8205
	Accurate	0.8485		0.866		0.8571		0.8079		0.8205
BLSTM Mazajak Skip-Gram
	Inaccurate	0.8542		0.6949		0.7664		0.8222		0.8397
	Accurate	0.8333		0.9278^b		0.8780		0.8222		0.8397
BLSTM ArWordVec Skip-Gram
	Inaccurate	0.8261		0.6441		0.7238		0.7919		0.8141
	Accurate	0.8091		0.9175		0.8148		0.7919		0.8141
BLSTM ArWordVec CBOW
	Inaccurate	0.7925		0.7119		0.75		0.805		0.8205
	Accurate	0.835		0.8866		0.86		0.805		0.8205
BLSTM AraVec CBOW
	Inaccurate	0.6865		0.7797		0.7302		0.7737		0.7821
	Accurate	0.8571		0.866		0.8172		0.7737		0.7821
BLSTM AraVec Skip-Gram
	Inaccurate	0.7313		0.8305^d		0.7777		0.8136		0.8205
	Accurate	0.8144		0.8144		0.8494		0.8136		0.8205
BLSTM fastText
	Inaccurate	0.8158		0.5254		0.6392		0.7382		0.7756
	Accurate	0.7627		0.9278^b		0.8372		0.7382		0.7756

^aAraBERTv2: Transformer-based Model for Arabic Language Understanding version 2.

^bRepresents the second-best value.

^cAraBERTv0.2: Transformer-based Model for Arabic Language Understanding version 0.2.

^dIndicates the best value.

^eBERT: bidirectional encoder representations from transformers.

^fBLSTM: bidirectional long short-term memory.

^gCBOW: Continuous Bag of Words.