Skip to main content

Table 2 Accuracy on aspect term level ABSA on four English datasets

From: Attention-based Sentiment Reasoner for aspect-based sentiment analysis

Model Res (2014) Lap (2014) Res (2015) Res (2016)
LSTM 74.28 66.45
TD-LSTM 75.63 68.13
ATAE-LSTM 77.20 68.70
MemNN 78.38 71.11 73.36 84.15
IAN 78.60 71.78 77.55 85.84
RAM 78.93 71.81 78.56 87.08
AS-Reasoner-I 79.73 72.41 79.06 85.69
AS-Reasoner-G 79.64 72.57 78.56 86.62
AS-Reasoner 80.00 74.77 79.90 86.31
  1. Italic values indicate the significance of the best result
  2. “Res” denotes the restaurant domain, “Lap” denotes the laptop domain