TTC Web Platform: from Corpus Compilation to Bilingual Terminologies for MT and CAT Tools
Résumé
This paper describes the TTC Web platform, an online demonstrator to show the whole pipeline to compile bilingual terminologies out of comparable corpora gathered from the web using the tools developed in the TTC project Terminology Extraction, Translation Tools and Comparable Corpora. We present the whole chain which has been integrated into the platform, as well as their main components: a focused web crawler; a UIMA based tool for both monolingual term extraction and bilingual term alignment, tools for monolingual term extraction using both rule-based and probabilistic methods, and finally, an online terminology platform to edit the output of the TTC tools. The TTC tool chain is available for all the languages of the project: DE, EN, ES, FR, LV, RU and ZH. With respect to the potential users of the tools, in the first Tralogy conference we presented the different users and scenarios that were envisaged: from basic users to professionals of the MT industry. In this paper we will include the first feedback obtained from users during the second user workshop that was organized to demonstrate and test the tools with potential users and experts of the MT, CAT, and terminology management domain.
Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...