What’s Hard in English RST Parsing? Predictive Models for Error Analysis

Abstract

Despite recent advances in Natural Language Processing (NLP), hierarchical discourse parsing in the framework of Rhetorical Structure Theory remains challenging, and our understanding of the reasons for this are as yet limited. In this paper, we examine and model some of the factors associated with parsing difficulties in previous work: the existence of implicit discourse relations, challenges in identifying long-distance relations, out-of-vocabulary items, and more. In order to assess the relative importance of these variables, we also release two annotated English test-sets with explicit correct and distracting discourse markers associated with gold standard RST relations. Our results show that as in shallow discourse parsing, the explicit/implicit distinction plays a role, but that long-distance dependencies are the main challenge, while lack of lexical overlap is less of a problem, at least for in-domain parsing. Our final model is able to predict where errors will occur with an accuracy of 76.3% for the bottom-up parser and 76.6% for the top-down parser.

Theoretical Studies of Natural Gas Hydrates and H-bonded Clusters and Crystals

“…The results above only show the correlation of standalone factors with label shift, without considering all factors together. Inspired by Liu et al (2023c), we train an XGBoost model (Chen and Guestrin, 2016) to find out the importance of factors when using the four features to predict the calculated label shift metric. XGBoost is a gradient boosting framework, where the importance of a feature can be measured by the performance gain it brings (Shang et al, 2019).…”