Comparison by Conversion: Reverse-Engineering UCCA from Syntax and Lexical Semantics

Proceedings of the 28th International Conference on Computational Linguistics, pages 2947–2966 Barcelona, Spain (Online), December 8-13, 2020 2947 Comparison by Conversion: Reverse-Engineering UCCA from Syntax and Lexical Semantics Daniel Hershcovich♦ Nathan Schneider♠ Dotan Dvir♥ Jakob Prange♠ Miryam de Lhoneux♦ Omri Abend♥ ♦University of Copenhagen ♠Georgetown University ♥Hebrew University of Jerusalem {dh,ml}@di.ku.dk, {nathan.schneider,jp1724}@georgetown.edu, dotan.dvir@mail.huji.ac.il, oabend@cs.huji.ac.il Abstract Building robust natural language understanding systems will require a clear characterization of whether and how various linguistic meaning representations complement each other. To per- form a systematic comparative analysis, we evaluate the mapping between meaning representa- tions from different frameworks using two complementary methods: (i) a rule-based converter, and (ii) a supervised delexicalized parser that parses to one framework using only information from the other as features. We apply these methods to convert the STREUSLE corpus (with syntactic and lexical semantic annotations) to UCCA (a graph-structured full-sentence mean- ing representation). Both methods yield surprisingly accurate target representations, close to fully supervised UCCA parser quality—indicating that UCCA annotations are partially redun- dant with STREUSLE annotations. Despite this substantial convergence between frameworks, we find several important areas of divergence. 1 Comparing Meaning Representations Several symbolic meaning representations (MRs) support human annotation of text with broad coverage (Abend and Rappoport, 2017; Oepen et al., 2019). To date, it is still not completely clear, for all frame- works, what linguistic semantic phenomena they encode, and how it compares to the content represented by the others. It therefore behooves us to develop a firm linguistic understanding of MRs. In particular: are they merely a coarsening and rearranging of syntactic information, such as is encoded in Universal Dependencies (UD; Nivre et al., 2016, 2020)? To what extent do they take lexical semantic proper- ties into account? What does this suggest about the potential for exploiting simpler or better-resourced linguistic representations for improved MR parsing? Intuitively, we ask whether: sentence-level MR ?= syntax + lexical semantics To address this question, we examine UCCA, a document-level MR often used for sentence-level semantics (see §2.1). Hershcovich et al. (2019) began to examine the relation of UCCA to syntax, con- tributing a corpus with gold standard UD and UCCA parses, heuristically aligning them, and quantifying the correlations between syntactic and semantic labels. Conversely, Hershcovich et al. (2018) provided some initial evidence that other MRs can be brought to bear on the UCCA parsing task via multitask learning, but left the details of the relationship between representations to latent (and opaque) parame- ters of neural models. In this paper, we aim to close the gap between the two previous investigations by (1) building an interpretable rule-based system to convert from shallower representations (syntax and lexical semantic units/tags) into UCCA, forcing us to be linguistically precise about what UCCA captures and how it “decomposes”; and (2) training top-performing supervised parsers in a delexicalized setting with only syntactic and lexical semantic features, as a data-driven mapping corroborating the rule-based approach. We perform our analysis on the Reviews section of the English Web Treebank (Bies et al., 2012), which …

Content truncated to 500 words