RST Parsing from Scratch
This work addresses discourse parsing for natural language processing applications, offering a faster and more adaptable approach without handcrafted features, though it is incremental in advancing existing parsing methods.
The authors tackled document-level discourse parsing in the Rhetorical Structure Theory framework by introducing a novel top-down end-to-end formulation that treats parsing as a sequence of splitting decisions, eliminating the need for prerequisite discourse segmentation. Their parser outperformed existing methods on the standard English RST discourse treebank, achieving better results in both end-to-end parsing and parsing with gold segmentation.
We introduce a novel top-down end-to-end formulation of document-level discourse parsing in the Rhetorical Structure Theory (RST) framework. In this formulation, we consider discourse parsing as a sequence of splitting decisions at token boundaries and use a seq2seq network to model the splitting decisions. Our framework facilitates discourse parsing from scratch without requiring discourse segmentation as a prerequisite; rather, it yields segmentation as part of the parsing process. Our unified parsing model adopts a beam search to decode the best tree structure by searching through a space of high-scoring trees. With extensive experiments on the standard English RST discourse treebank, we demonstrate that our parser outperforms existing methods by a good margin in both end-to-end parsing and parsing with gold segmentation. More importantly, it does so without using any handcrafted features, making it faster and easily adaptable to new languages and domains.