Bracketing and aligning words and constituents in parallel text using Stochastic Inversion Transduction Grammars

Abstract

Abstract:We introduce (1) a novel stochastic inversion transduction grammar formalism for bilingual language modeling of sentence-pairs, and (2) the concept of bilingual parsing with a variety of parallel corpus analysis applications. Aside from the bilingual orientation, three major features distinguish the formalism from the finitestate transducers more traditionally found in computational linguistics: it skips directly to a context-free rather than finite-state base, it permits a minimal extra degree of ordering flexibility, and its probabilistic formulation admits an efficient maximum-likelihood bilingual parsing algorithm. A convenient normal form is shown to exist. Analysis of the formalism's expressiveness suggests that it is particularly well-suited to model ordering shifts between languages, balancing needed flexibility against complexity constraints. We discuss a number of examples of how stochastic inversion transduction grammars bring bilingual constraints to bear upon problematic corpus analysis tasks such as segmentation, bracketing, phrasal alignment, and parsing.

Unsupervised Generation of Parallel Treebanks through Sub-Tree Alignment

“…In its operation, the string-to-string aligner is very similar to ITG (Wu, 2000), however its goal is the generation of a parallel treebank, rather than the induction of a bilingual grammar.…”

Semantic Frames as Interlingual Representations for Multilingual Lexical Databases

“…This procedure is facilitated by the use of parallel-aligned corpora, which allow a comparison between the LUs when they are embedded in different types of context (see, e.g. Wu 2000, Salkie 2002). 14 Consider, for example, the verb answer, whose individual frame elements may be realized syntactically in many different ways.…”

“…A first step is to use parallel corpora to automatically identify translation equivalents in context in order to determine frame membership of lexical units across languages. For approaches incorporating automatic acquisition of lexical information from parallel corpora see Wu (2000), Farwell et al (2004), Green et al (2004), and Mitamura et al (2004).…”

Using bilingual dependencies to align words in Enlish/French parallel corpora

“…There are basically two kinds of systems working at these segmentation levels: the most widespread rely on statistical models, in particular the IBM ones (Brown et al, 1993); others combine simpler association measures with different kinds of linguistic information (Arhenberg et al, 2000;Barbu, 2004). Mainly dedicated to machine translation, purely statistical systems have gradually been enriched with syntactic knowledge (Wu, 2000;Yamada & Knight, 2001;Ding et al, 2003;Lin & Cherry, 2003). As pointed out in these studies, the introduction of linguistic knowledge leads to a significant improvement in alignment quality.…”