Transfer-Driven Machine Translation
نویسنده
چکیده
Transfer-Driven Machine Translation (TDMT) [1, 2] is a translation technique developed as a research project at ATR Interpreting Telecommunications Research Laboratories. In TDMT, translation is performed mainly by a transfer module which applies transfer knowledge to an input sentence. Other modules, such as lexical processing, analysis, contextual processing and generation, cooperate with the transfer module to improve translation performance. This transfer-centered mechanism can achieve efficient and robust translation by making the most of the example-based framework, which calculates a semantic distance between linguistic expressions. A TDMT prototype system is written in LISP and is demonstrated on a SUN workstation. In our TDMT demonstration, the following items are presented.
منابع مشابه
A Similarity-Driven Transfer System
The transfer phase in machine translation (MT) systems has been considered to be more complicated titan analysis and generation~ since it is inherently a conglomeration of individual lexical rules. Currently some at tempts are being made to use case-based reasoning in machine translation, that is, to make decisions on the basis of translation examples at appropriate pohtts in MT. This paper pro...
متن کاملCorpus-Driven Bilingual Lexicon Extraction
This paper introduces some key aspects of machine translation in order to situate the role of the bilingual lexicon in transfer-based systems. It then discusses the data-driven approach to extracting bilingual knowledge automatically from bilingual texts, tracing the processes of alignment at different levels of granularity. The paper concludes with some suggestions for future work. 1 Machine T...
متن کاملImproving Performance of Transfer-Driven Machine Translation with Extra-Linguistic Informatioon from Context, Situation and Environment
This paper describes an improvement in the performance of Transfer-Driven Machine Translation ( T D M T ) by the use of extralinguistic information. In evaluating what constitutes natural speech, particular attention was paid to word usage that depended on the extra-linguistic information from the context, situation, environment and so on. We discuss what types of extra-linguistic information a...
متن کاملAn Example-Based Method for Transfer-Driven Machine Translation
This paper presents a method called Transfer-Driven Machine Translation (TDMT), which utilizes an example-based framework for various process and combines multi-level knowledge. An example-based framework can achieve quick processing and consistently describe knowledge. It is useful for spoken-language translation, which needs robust and efficient translation. TDMT strengthens the example-based...
متن کاملTowards A Universal Tool For NLP Resource Acquisition
This paper describes an approach to developing a universal tool for eliciting, from a non-expert human user, knowledge about any language L. The purpose of this elicitation is rapid development of NLP systems. The approach is described on the example of the syntax module of the Boas knowledge elicitation system for a quick ramp up of a standard transfer-based machine translation system from L i...
متن کامل