Abstract
AbstractLexicalized parsing models are based on the assumptions that (i) constituents are organized around a lexical head and (ii) bilexical statistics are crucial to solve ambiguities. In this paper, we introduce an unlexicalized transition-based parser for discontinuous constituency structures, based on a structure-label transition system and a bi-LSTM scoring system. We compare it with lexicalized parsing models in order to address the question of lexicalization in the context of discontinuous constituency parsing. Our experiments show that unlexicalized models systematically achieve higher results than lexicalized models, and provide additional empirical evidence that lexicalization is not necessary to achieve strong parsing results. Our best unlexicalized model sets a new state of the art on English and German discontinuous constituency treebanks. We further provide a per-phenomenon analysis of its errors on discontinuous constituents.
Highlights
This paper introduces an unlexicalized parsing model and addresses the question of lexicalization, as a parser design choice, in the context of transition-based discontinuous constituency parsing
In a lexicalized Probabilistic ContextFree Grammar (PCFG), grammar rules involve nonterminals annotated with a terminal element that represents their lexical head, for example: VP[saw] −→ VP[saw] PP[telescope]
Bikel (2004) showed that bilexical statistics were rarely used during decoding, and that when used, they were close to that of backoff distributions used for unknown word pairs
Summary
This paper introduces an unlexicalized parsing model and addresses the question of lexicalization, as a parser design choice, in the context of transition-based discontinuous constituency parsing. Lexicalized parsing models (Collins, 1997; Charniak, 1997) are based on the assumptions that (i) constituents are organized around a lexical head and (ii) bilexical statistics are crucial to solve ambiguities. In a lexicalized Probabilistic ContextFree Grammar (PCFG), grammar rules involve nonterminals annotated with a terminal element that represents their lexical head, for example: VP[saw] −→ VP[saw] PP[telescope]. The probability of such a rule models the likelihood that telescope is a suitable modifier for saw. Bikel (2004) showed that bilexical statistics were rarely used during decoding, and that when used, they were close to that of backoff distributions used for unknown word pairs
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have