AT-LANG
The distribution of language resources in an organisation across different systems makes reusability difficult. AT-LANG centralises dictionaries, translation memories and grammatical rules for different uses. This way, maintaining terminological coherence and standardisation is made easier.
The language pairs available are:
- Spanish with Basque, Catalan, Galician, Portuguese (European and of Brazil), French and English.
- Catalan with Spanish, Basque, Galician, Portuguese (European and of Brazil), French and English.
- Galician with Spanish, Basque, Catalan, Portuguese (European and of Brazil), French and English.
Each language pair includes: rigorously tagged general and specific dictionaries; static and dynamic translation memories; and a system of rules.
Language resources are stored in an optimised repository for the use of the AT-CORE and AT-COR , linguistic processors that are executed at the maximum speed to effectively carry out the necessary steps for each translation and correction.
The general and specific dictionaries for each language, and the translation memory with set phrases and sayings are categorised with morphological and syntactic, as well as command and disambiguation labels in order to be able to discern the most appropriate translation and correction in each case.
Analysis and generation rules refer to those which determine the placement of words in relation to grammar and the correct collocations in each translation.
The combination of these three elements (dictionaries, translation memories and grammatical rules) leads to a high-quality translation with an error rate of less than 2% in well-trained language pairs.
