Treebank-Based Acquisition of Multilingual Unification Grammar Resources

Abstract

Abstract. Deep unification-(constraint-)based grammars are usually hand-crafted. Scaling such grammars from fragments to unrestricted text is time-consuming and expensive. This problem can be exacerbated in multilingual broad-coverage grammar development scenarios. Cahill et al. (2002Cahill et al. ( , 2004 and O'Donovan et al. (2004) present an automatic f-structure annotation-based methodology to acquire broad-coverage, deep, Lexical-Functional Grammar (LFG) resources for English from the Penn-II Treebank. In this paper we show how this model can be adapted to a multilingual grammar development scenario to induce robust, wide-coverage, PCFG-based LFG approximations for German from the TIGER Treebank. We show how the architecture of LFG, in particular the distinction between c-structure and f-structure representations, facilitates multilingual, treebank-based unification grammar induction, allowing us to cross-linguistically reuse the lexical extraction and parsing modules from O' Donovan et al. (2004) and Cahill et al. (2004), respectively. We evaluate our grammars against the PARC 700 Dependency Bank , against dependency structures for 2000 held-out sentences from the TIGER Corpus as well as against a hand-crafted dependency gold standard for 100 TIGER trees. Currently, our resources achieve 81.79% f-score against the PARC 700, a 2.19% improvement over the best result reported for a hand-crafted grammar in Kaplan et al. (2004), 74.6% against the 2000 held-out TIGER dependency structures and 71.08% against the 100-sentence TIGER gold standard, with substantially improved coverage compared to hand-crafted resources. We have since applied our methodology to induce wide-coverage LFG resources for Chinese (Burke et al., 2004b) from the Penn Chinese Treebank (Xue et al., 2002) and for Spanish from the CAST3LB Treebank (Civit, 2003).

Urdu in a parallel grammar development environment

“…Evaluation of the ParGram LFG grammars has focused on accuracy measures against industry-determined standards such as the Penn Treebank for English and the Tiger Treebank for German. To evaluate against these resources, dependency banks are semi-automatically built for the treebanks (see Cahill et al (2005) and references therein for a general approach and Forst (2003a, b) and Forst et al (2004) on German). In addition, gold standard dependency banks, like the for English , have been built for some languages.…”

Dublin City University at QA@CLEF 2008

“…Finally, the answer selection component, which matches questions to the answers, consists of passage retrieval, sentence analysis, and answer reranking components. In the current system, we used TreeTagger for POS tagging and Chunking, and a treebank-based Lexical Functional Grammar(LFG) parser for dependecy parsing [2,14].…”

“…For this, we parsed both the questions and the sentences using a Lexical Functional Grammar(LFG)-based parser [2] developed at Dublin City University. The system takes the output of a syntactic parser (Charniak parser [3]) and generates an F-Structure, a labeled bilexical dependency graph.…”

A Parallel Metagrammar for Closely Related Languages – A Case Study of Czech and Russian

“…Treebank-based methodology, as we use it, is different from that in Cahill et al (2003) where wide-coverage probabilistic LFGs for English and German are automatically obtained from Penn-II and TIGER corpora. Our metagrammar for Czech was prepared manually.…”