W-RST: Towards a Weighted RST-style Discourse Framework
This work addresses a specific issue in NLP for discourse analysis, offering an incremental improvement by integrating data-driven and linguistic approaches.
The paper tackles the problem of binary importance assessment in discourse analysis by proposing a Weighted-RST framework that uses real-valued scores, showing it benefits NLP downstream applications compared to nuclearity-centered approaches and partially aligns with human annotator assessments.
Aiming for a better integration of data-driven and linguistically-inspired approaches, we explore whether RST Nuclearity, assigning a binary assessment of importance between text segments, can be replaced by automatically generated, real-valued scores, in what we call a Weighted-RST framework. In particular, we find that weighted discourse trees from auxiliary tasks can benefit key NLP downstream applications, compared to nuclearity-centered approaches. We further show that real-valued importance distributions partially and interestingly align with the assessment and uncertainty of human annotators.