The theory of discourse functional pragmatics posits that every discourse unit serves a purpose, and functional pragmatics can reveal its role in the text. To achieve a more informative representation of paragraphs, m...
详细信息
ISBN:
(纸本)9789819756711;9789819756728
The theory of discourse functional pragmatics posits that every discourse unit serves a purpose, and functional pragmatics can reveal its role in the text. To achieve a more informative representation of paragraphs, mitigate the long-tail problem, and incorporate the specificity of different models, we propose a Chinese discourse Functional Pragmatics Recognition model based on Multi-level information and Ensemble learning (FPRME). Specifically, two different encoding methods are combined to obtain paragraph representation that contains rich word-level information. Then, an encoder module is used to perform paragraph interactions, and graph convolutional networks are applied to enhance the interaction between paragraphs with dependencies. After that, a deep differential amplifier is applied to amplify the difference between paragraphs, and a weighted loss function is used to balance the attention of the categories. In addition, text augmentation and adversarial training are also employed to enhance the focus on minorities. Finally, ensemble learning is used to further improve the model's performance. The experimental results on MCDTB 2.0 demonstrate that our model FPRME outperforms the state-of-the-art models.
暂无评论