Attentional Encoder Network for Targeted Sentiment Classification (arXiv 2019)

- 5 mins

Abstract

1. Introduction

3. Proposed Methodology

figure1

3.1 Embedding Layer

3.1.1 GloVe Embedding

3.1.2 BERT Embedding

3.2 Attentional Encoder Layer

3.2.1 Multi-Head Attention

3.2.2 Point-wise Convolution Transformation

3.3 Target-specific Attention Layer

3.4 Output Layer

3.5 Regularization and Model Training

eq151617

4. Experiments

4.1 Dataset and Experimental Settings

4.2 Model Comparisons

4.3 Main Results

table2

4.4 Model Analysis

table3

5. Conclusion

Joohong Lee

Joohong Lee

Machine Learning Researcher

rss facebook twitter github youtube mail spotify instagram linkedin google pinterest medium