ImpAr: A Deterministic Algorithm for Implicit Semantic Role Labelling

Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pages 1180–1189, Sofia, Bulgaria, August 4-9 2013. c⃝2013 Association for Computational Linguistics ImpAr: A Deterministic Algorithm for Implicit Semantic Role Labelling Egoitz Laparra IXA Group University of the Basque Country San Sebastian, Spain egoitz.laparra@ehu.es German Rigau IXA Group University of the Basque Country San Sebastian, Spain german.rigau@ehu.es Abstract This paper presents a novel deterministic algorithm for implicit Semantic Role La- beling. The system exploits a very sim- ple but relevant discursive property, the ar- gument coherence over different instances of a predicate. The algorithm solves the implicit arguments sequentially, exploit- ing not only explicit but also the implicit arguments previously solved. In addition, we empirically demonstrate that the algo- rithm obtains very competitive and robust performances with respect to supervised approaches that require large amounts of costly training data. 1 Introduction Traditionally, Semantic Role Labeling (SRL) sys- tems have focused in searching the fillers of those explicit roles appearing within sentence bound- aries (Gildea and Jurafsky, 2000, 2002; Carreras and M`arquez, 2005; Surdeanu et al., 2008; Hajiˇc et al., 2009). These systems limited their search- space to the elements that share a syntactical re- lation with the predicate. However, when the par- ticipants of a predicate are implicit this approach obtains incomplete predicative structures with null arguments. The following example includes the gold-standard annotations for a traditional SRL process: (1) [arg0 The network] had been expected to have [np losses] [arg1 of as much as $20 million] [arg3 on base- ball this year]. It isn’t clear how much those [np losses] may widen because of the short Series. The previous analysis includes annotations for the nominal predicate loss based on the NomBank structure (Meyers et al., 2004). In this case the annotator identifies, in the first sentence, the argu- ments arg0, the entity losing something, arg1, the thing lost, and arg3, the source of that loss. How- ever, in the second sentence there is another in- stance of the same predicate, loss, but in this case no argument has been associated with it. Tradi- tional SRL systems facing this type of examples are not able to fill the arguments of a predicate because their fillers are not in the same sentence of the predicate. Moreover, these systems also let unfilled arguments occurring in the same sentence, like in the following example: (2) Quest Medical Inc said it adopted [arg1 a sharehold- ers’ rights] [np plan] in which rights to purchase shares of common stock will be distributed as a dividend to shareholders of record as of Oct 23. For the predicate plan in the previous sentence, a traditional SRL process only returns the filler for the argument arg1, the theme of the plan. However, in both examples, a reader could eas- ily infer the missing arguments from the surround- ing context of the predicate, and determine that in (1) both instances of the predicate share the same arguments and in (2) the missing argument corresponds to the subject …

Content truncated to 500 words