Semantic Role Labeling as Syntactic Dependency Parsing

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 7551–7571, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 7551 Semantic Role Labeling as Syntactic Dependency Parsing Tianze Shi∗ Cornell University tianze@cs.cornell.edu Igor Malioutov Bloomberg L.P. imalioutov@bloomberg.net Ozan ˙Irsoy Bloomberg L.P. oirsoy@bloomberg.net Abstract We reduce the task of (span-based) PropBank- style semantic role labeling (SRL) to syntac- tic dependency parsing. Our approach is mo- tivated by our empirical analysis that shows three common syntactic patterns account for over 98% of the SRL annotations for both En- glish and Chinese data. Based on this obser- vation, we present a conversion scheme that packs SRL annotations into dependency tree representations through joint labels that permit highly accurate recovery back to the original format. This representation allows us to train statistical dependency parsers to tackle SRL and achieve competitive performance with the current state of the art. Our findings show the promise of syntactic dependency trees in en- coding semantic role relations within their syn- tactic domain of locality, and point to potential further integration of syntactic methods into semantic role labeling in the future. 1 Introduction Semantic role labeling (SRL; Palmer et al., 2010) analyzes texts with respect to predicate argument structures such as “who did what to whom, and how, when and where”. These generic surface semantic representations provide richer linguistic analysis than syntactic parsing alone and are useful in a wide range of downstream applications including ques- tion answering (Shen and Lapata, 2007; Khashabi et al., 2018), open-domain information extraction (Christensen et al., 2010), clinical narrative under- standing (Albright et al., 2013), automatic summa- rization (Khan et al., 2015) and machine translation (Liu and Gildea, 2010; Xiong et al., 2012; Bazraf- shan and Gildea, 2013), among others. It is commonly acknowledged that syntax and se- mantics are tightly coupled with each other (Levin ∗Work done during an internship at Bloomberg L.P. She wanted to design the bridge ... nsubj-A0- - xcomp-A1-(A0,A0)- mark- - - dobj-A1- - det- - - A0 pred A1 A0 A1 pred Figure 1: An example sentence with SRL annotations (below) and our joint syntacto-semantic dependency re- lations (above; described in §3). The two representa- tions can be converted from one to the other. A0 and A1 are short for SRL relations ARG0 and ARG1. and Hovav, 2005). In some forms of linguistic the- ories (Baker, 1996, 1997), semantic arguments are even hypothesized to be assigned under consistent and specific syntactic configurations. As a matter of practice, annotations of semantic roles (Palmer et al., 2005, inter alia) are typically based on exist- ing syntactic treebanks as an additional annotation layer. Annotators are instructed (Babko-Malaya et al., 2006; Bonial et al., 2015) to identify seman- tic arguments within the predicates’ domain of lo- cality,1 respecting the strong connection between syntax and semantics. Empirically, syntax has indeed been shown to be helpful to SRL in a variety of ways. Earlier SRL systems have successfully incorporated syn- tactic parse trees as features and pruning signals (Punyakanok et …

Content truncated to 500 words