Top-down Discourse Parsing via Sequence Labelling
This work provides a conceptually simpler and more efficient method for discourse parsing, which is beneficial for natural language understanding tasks for NLP researchers.
This paper tackles discourse parsing by reframing it as a sequence labeling problem to iteratively segment documents into discourse units. This approach eliminates the need for a decoder and reduces the search space, with their LSTM model achieving a new state-of-the-art for RST parsing based on the Full metric.
We introduce a top-down approach to discourse parsing that is conceptually simpler than its predecessors (Kobayashi et al., 2020; Zhang et al., 2020). By framing the task as a sequence labelling problem where the goal is to iteratively segment a document into individual discourse units, we are able to eliminate the decoder and reduce the search space for splitting points. We explore both traditional recurrent models and modern pre-trained transformer models for the task, and additionally introduce a novel dynamic oracle for top-down parsing. Based on the Full metric, our proposed LSTM model sets a new state-of-the-art for RST parsing.