MIT Press

Skip to Main Content

Table 6:

Performance comparison of competing models in the Yelp 2013 Sparse80 data set.

Models	Accuracy
NSC	51.1
BiLSTM +CSAA	52.7
HCSC	53.8
BiLSTM +encoder-basis-cust	50.4
BiLSTM +linear-basis-cust	50.8
BiLSTM +bias-basis-cust	51.9
BiLSTM +word-basis-cust	51.9
BiLSTM +attention-basis-cust	53.1

Models	Accuracy
NSC	51.1
BiLSTM +CSAA	52.7
HCSC	53.8
BiLSTM +encoder-basis-cust	50.4
BiLSTM +linear-basis-cust	50.8
BiLSTM +bias-basis-cust	51.9
BiLSTM +word-basis-cust	51.9
BiLSTM +attention-basis-cust	53.1