Transition-Based Techniques for Non-Projective Dependency Parsing

Abstract

We present an empirical evaluation of three methods for the treatment of non-projective structures in transition-based dependency parsing: pseudo-projective parsing, non-adjacent arc transitions, and online reordering. We compare both the theoretical coverage and the empirical performance of these methods using data from Czech, English and German. The results show that although online reordering is the only method with complete theoretical coverage, all three techniques exhibit high precision but somewhat lower recall on non-projective dependencies and can all improve overall parsing accuracy provided that non-projective dependencies are frequent enough. We also find that the use of non-adjacent arc transitions may lead to a drop in accuracy on projective dependencies in the presence of long-distance non-projective dependencies, an effect that is not found for the two other techniques.

A Linear-Time Transition System for Crossing Interval Trees

“…We choose the swap-based system as our non-projective baseline as it currently represents the state-of-the-art in transition-based parsing (Bohnet et al, 2013), with higher empirical performance than the Attardi system or pseudo-projective parsing (Kuhlmann and Nivre, 2010).…”

“…The experiments compare the two-registers transition system for mildly non-projective trees proposed here with two other transition systems: the arceager system for projective trees (Nivre, 2003) and the swap-based system for all non-projective trees (Nivre, 2009). We choose the swap-based system as our non-projective baseline as it currently represents the state-of-the-art in transition-based parsing (Bohnet et al, 2013), with higher empirical performance than the Attardi system or pseudo-projective parsing (Kuhlmann and Nivre, 2010).…”

Grammatical Case Based IS-A Relation Extraction with Boosting for Polish

“…It is shown in [15] and [16] that dependency parsing for non-projective grammars is NP-hard, apart from a very narrow subclass called edgefactored grammars. This challenge is addressed, among others, by transition-based dependency parsing [17] used in the preprocessing step for the algorithm described in this paper. We argue that inflection in a language is not only a drawback but (-mi) can also be a great advantage.…”

Word Ordering as a Graph Rewriting Process

“…Formal DGs handling non-projective ordering have been proposed [10,11,12,13], but they do not fit our first aim, which is to use the same formalism for linearization and subcategorization rules. Non-projective dependency parsers have been proposed [14], but the grammar cannot be clearly separated from the parsing procedure. Meaning-Text Theory (MTT) [15] is a model we want to follow because it is very modular and all rules are expressed in similar terms, that is, in terms of correspondence between two levels of representation.…”