Improving Neural RST Parsing Model with Silver Agreement Subtrees

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1600–1612 June 6–11, 2021. ©2021 Association for Computational Linguistics 1600 Improving Neural RST Parsing Model with Silver Agreement Subtrees Naoki Kobayashi1, Tsutomu Hirao2, Hidetaka Kamigaito1, Manabu Okumura1, Masaaki Nagata2 1Institute of Innovative Research, Tokyo Institute of Technology, 2NTT Communication Science Laboratories, NTT Corporation {kobayasi@lr., kamigaito@lr., oku@}pi.titech.ac.jp {tsutomu.hirao.kp, masaaki.nagata.et}@hco.ntt.co.jp Abstract Most of the previous Rhetorical Structure The- ory (RST) parsing methods are based on su- pervised learning such as neural networks, that require an annotated corpus of sufficient size and quality. However, the RST Discourse Treebank (RST-DT), the benchmark corpus for RST parsing in English, is small due to the costly annotation of RST trees. The lack of large annotated training data causes poor performance especially in relation labeling. Therefore, we propose a method for improv- ing neural RST parsing models by exploiting silver data, i.e., automatically annotated data. We create large-scale silver data from an un- labeled corpus by using a state-of-the-art RST parser. To obtain high-quality silver data, we extract agreement subtrees from RST trees for documents built using the RST parsers. We then pre-train a neural RST parser with the obtained silver data and fine-tune it on the RST-DT. Experimental results show that our method achieved the best micro-F1 scores for Nuclearity and Relation at 75.0 and 63.2, re- spectively. Furthermore, we obtained a re- markable gain in the Relation score, 3.0 points, against the previous state-of-the-art parser. 1 Introduction Rhetorical Structure Theory (RST) (Mann and Thompson, 1987) is one of the most widely used theories for representing the discourse structure of a text as a tree. RST trees are a kind of con- stituent tree, whose leaves are Elementary Dis- course Units (EDUs), i.e., clause-like units, and whose non-terminal nodes cover text spans con- sisting of either a sequence of EDUs or a single EDU. The label of a non-terminal node represents the attribution of a text span, i.e., nucleus (N) or satellite (S). A discourse relation is also assigned between two adjacent non-terminal nodes. In most cases, RST parsers have been devel- oped on the basis of supervised learning algorithms (Wang et al., 2017b; Yu et al., 2018; Kobayashi et al., 2020; Lin et al., 2019; Zhang et al., 2020), which require a high-quality annotated corpus of sufficient size. Generally, they train the following three components of the RST parsing: (1) struc- ture prediction by splitting a text span consisting of contiguous EDUs into two smaller ones or merg- ing two adjacent spans into a larger one, (2) nu- clearity status prediction for two adjacent spans by solving a 3-class classification problem, and (3) relation label prediction for two adjacent spans by solving an 18-class classification problem (see Section 3.3 for details). However, it is costly to annotate RST trees for a huge collection of doc- uments, and thus it is difficult to obtain a large amount of human-annotated data for RST parsing. As a result, …

Content truncated to 500 words