Attentional Encoder Network for Targeted Sentiment Classification
This addresses sentiment analysis for specific targets in text, offering an incremental improvement over existing methods.
The paper tackles targeted sentiment classification by proposing an Attentional Encoder Network (AEN) that replaces RNNs with attention-based encoders to improve parallelization and long-term pattern retention, achieving new state-of-the-art results with pre-trained BERT.
Targeted sentiment classification aims at determining the sentimental tendency towards specific targets. Most of the previous approaches model context and target words with RNN and attention. However, RNNs are difficult to parallelize and truncated backpropagation through time brings difficulty in remembering long-term patterns. To address this issue, this paper proposes an Attentional Encoder Network (AEN) which eschews recurrence and employs attention based encoders for the modeling between context and target. We raise the label unreliability issue and introduce label smoothing regularization. We also apply pre-trained BERT to this task and obtain new state-of-the-art results. Experiments and analysis demonstrate the effectiveness and lightweight of our model.