Integrating Knowledge Bases and Statistics in MT

TL;DR

This paper discusses combining knowledge-based and statistical methods to improve machine translation, specifically focusing on scaling up grammar-based techniques for Japanese-English newspaper translation.

Contribution

It introduces a hybrid approach that integrates knowledge bases with statistical methods to enhance the scalability and effectiveness of grammar-based MT systems.

Findings

Successful integration of statistical methods with knowledge-based MT

Improved translation quality for Japanese-English newspaper texts

Enhanced scalability of grammar-based MT techniques

Abstract

We summarize recent machine translation (MT) research at the Information Sciences Institute of USC, and we describe its application to the development of a Japanese-English newspaper MT system. Our work aims at scaling up grammar-based, knowledge-based MT techniques. This scale-up involves the use of statistical methods, both in acquiring effective knowledge resources and in making reasonable linguistic choices in the face of knowledge gaps.