Description
Project Description
Objectives
The main objectives of TT2 are two-fold:
- to provide a framework in which leading-edge research can be conducted in the area of data driven methods in natural language processing and, more specifically, translation.
- to provide a practical application for that research which will help solve a pressing social problem: how to meet the ever-growing demand for high-quality translation.
The research conducted in our project will be applied to the development of an innovative, interactive machine translation system. TT2 will aim at facilitating the task of producing high-quality translations, and make the translation task more cost-effective for human translators. Research progress will thus be measurable in terms of the increased productivity of translators using this new computer-assisted translation system.
The proposed translation process will keep two major facts as general requirements:
- the human translator is the key component to producing high quality output.
- machines provide the means to speed up the translation task and make it more efficient.
Finding a solution that meets these two requirements is the key challenge to our project. TT2 has been devised as an optimal solution to this challenge.
Background
The results of two previous projects, TransType and EuTrans, serve as starting points. TransType developed a tool for speeding up and facilitating the work of translators by automatically suggesting translation completions. Suggestions are provided as follows: with each new character the translator enters, the system revises its calculations and provides new suggestions that are compatible with the translator's new input. If the system provides the right suggestion, the translator has only to accept it, thereby saving time in producing his or her translation. Otherwise, the translator ignores the system's suggestions and continues to type his or her intended translation.
EuTrans focused on the development of novel approaches for speech and text automated translation-using data driven methods. Building on the results of TransType and EuTrans, TT2 will extend the latter's capabilities in the following directions:
- Extend TransType to provide completions for a wider range of multi-word units and longer sentence fragments.
- Add two new languages to the existing language pair (English into French): English into Spanish and English into German. In addition, systems will be developed for the reverse language directions: French, Spanish, and German into English.
- Investigate the possibility of integrating declarative sources of linguistic knowledge, e.g. domain-specific translation glossaries or dictionaries, with statistical machine translation technology.
- Further expand the user environment and improve the ergonomics of the system so that TT2 can be put into actual use in a conventional production environment like that of a translation agency.
Approach
TT2 aims at developing new techniques to make the task of producing high-quality translations more cost-effective, while keeping quality high. Two facts are key to this purpose:
- human translators are key to guaranteeing high quality output
- machines are key to speeding up the translation process and making it more efficient
This project aims at developing a solution that meets these two requirements by properly combining the positive aspects of the MT and CAT paradigms in one single environment, TT2. In order to produce high-quality translations, TT2 will put the translator in a central position in the translation process. In order to optimise translator's productivity, TT2 will include a data driven MT engine to assist the translator with suggestions.