Syntactic Knowledge Embedding Network for Aspect-Based Sentiment Classification

Table 2

Main experiment results (%). Best results are in bold (accuracy and , larger values show better performance).

Models	Laptop		Restaurant		Twitter
	Accuracy		Accuracy		Accuracy
MemNet	70.33	64.09	78.16	65.83	68.50	66.91
TD-LSTM	71.80	68.42	78.00	66.70	66.62	64.0
IAN	72.10	—	78.60	—	—	—
RAW	74.49	71.35	80.23	70.80	69.36	67.30
ASGCN	74.14	69.24	80.86	72.19	71.53	69.68
BERTbase	75.29	71.91	81.54	71.94	—	—
BERT-SPC	78.99	75.03	84.46	76.98	73.55	72.14
BERT-PT	78.08	75.08	84.95	76.96	—	—
BERT-SFE	79.47	75.90	85.18	78.15	75.43	72.75