Improve Discourse Dependency Parsing with Contextualized Representations
This improves discourse analysis for natural language processing applications, but it is incremental as it builds on existing work with transformers and sequence labeling.
The paper tackled discourse dependency parsing by proposing a method that uses transformers to encode contextualized representations for intra- and inter-sentential levels, treating relation identification as a sequence labeling task, and achieved state-of-the-art results on English and Chinese datasets.
Recent works show that discourse analysis benefits from modeling intra- and inter-sentential levels separately, where proper representations for text units of different granularities are desired to capture both the meaning of text units and their relations to the context. In this paper, we propose to take advantage of transformers to encode contextualized representations of units of different levels to dynamically capture the information required for discourse dependency analysis on intra- and inter-sentential levels. Motivated by the observation of writing patterns commonly shared across articles, we propose a novel method that treats discourse relation identification as a sequence labelling task, which takes advantage of structural information from the context of extracted discourse trees, and substantially outperforms traditional direct-classification methods. Experiments show that our model achieves state-of-the-art results on both English and Chinese datasets.