Abstract:For syntax-based tree-to-string statistical translation model, direct decoding algorithm is presented, which is based on the framework of the log-linear model. This algorithm can operates directly on the parsing trees, showing its effectiveness and convenience in exploiting syntactical information for the process of translation. For the log-linear model, three feature functions were defined, including the translation model probability, the language model probability and the null translation penalty. The direct decoding algorithm adopts beam search algorithm to probe the hypothesis space, uses multi-stacks to store incomplete hypotheses and applies heuristics to re-score these hypotheses. Among these techniques, the method of extending the hypothesis, especially the validity of nodes positions for different operations on the parsing tree, was carefully studied. Experiments were made on the IWSLT2004 data set. Experimental results show that the direct decoding algorithm gives better results than the conventional decoding algorithm according to the BLEU scores. This means that the direct decoding algorithm can overcome the shortages of the conventional decoding algorithm and makes more efficient use of syntactical information. Thus the direct decoding algorithm presented in this paper is more suitable for the syntax-based tree-to-string statistical translation model than the conventional decoding algorithm.

Some Issues of Syntax-based Statistical Machine Translation

Syntax-Based Chinese-Vietnamese Tree-to-Tree Statistical Machine Translation with Bilingual Features

A Tree-to-tree Alignment-Based Model for Statistical Machine Translation

Statistical Translation Model Based On Source Syntax Structure

Improving Phrase-Based Statistical Machine Translation Models by Incorporating Syntax-Based Language Models

A Statistical Machine Translation Model Based on a Synthetic Synchronous Grammar.

A Study of Translation Rule Classification for Syntax-Based Statistical Machine Translation

Filtering Syntactic Constraints for Statistical Machine Translation.

Subgraph based multitext grammar for statistical machine translation

Towards Integrated Machine Translation Using Structural Alignment From Syntax-Augmented Synchronous Parsing

Research of Tibetan syntax for machine translation

Dependency-based bracketing transduction grammar for statistical machine translation

Meta-Structure Transformation Model for Statistical Machine Translation.

Grammar Comparison Study for Translational Equivalence Modeling and Statistical Machine Translation.

A Source Dependency Model for Statistical Machine Translation.

Direct decoding algorithm for syntax-based tree-to-string statistical translation model

Syntax-aware Neural Machine Translation Directed by Syntactic Dependency Degree

Statistical Machine Translation for Speech: A Perspective on Structures, Learning, and Decoding

Linguistically motivated statistical machine translation : models and algorithms

An Improved Hierarchical Phrase Based Machine Translation Model

A Probabilistic Approach to Syntax-based Reordering for Statistical Machine Translation.