![]() Traitement des inconnus : une approche systématique de l’incomplétude lexicaleĪctes de la 17e conférence sur le Traitement Automatique des Langues Naturelles. At present, semantic lexica, a named entities guesser and a named entities phonetizer are being developed. The development of OAL has followed an incremental strategy. Today OAL manages resources in five European languages: French, English, Spanish, Italian and Polish. Moreover, different control mechanisms are set up to check the coherence and consistency of the resources. To add new words more easily to the morphosyntactic lexica, a guesser that lemmatizes and assigns morphosyntactic tags as well as inflection paradigms to a new word has been developed. ![]() In this paper we present the NLP architecture OAL, designed to assist computational linguists in the whole process of the development of resources in an industrial context: from corpora compilation to quality assurance. ![]() The creation, maintenance and enrichment of those resources are a labour-intensive task, especially when no tools are available. The performance of most NLP applications relies upon the quality of linguistic resources. Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10) OAL: A NLP Architecture to Improve the Development of Linguistic Resources for NLP LOL : Langage objet dédié à la programmation linguistique ( LOL: Object-oriented language dedicated to linguistic programming)Īctes de la 18e conférence sur le Traitement Automatique des Langues Naturelles. Proceedings of TextGraphs-6: Graph-based Methods for Natural Language Processingīabouk – exploration orientée du web pour la constitution de corpus et de terminologies (Babouk – oriented exploration of the web for the construction of corpora and terminologies)Īctes de la 18e conférence sur le Traitement Automatique des Langues Naturelles. SDMC has been developped in conjunction with the GREYC, MODYCO and LIPN institutes, using our sequential pattern mining algorithms.Grawl TCQ: Terminology and Corpora Building by Ranking Simultaneously Terms, Queries and Documents using Graph Random Walks SDMC provides a set of tools for sequential data mining. Sequential Data Mining under Constraints (SDMC) Grumph generates probabilistic phoneme lattices. Grumph is a grapheme-to-phoneme converter based on conditional random fields (CRFs) and weighted finite state transducers (WFSTs). HTML5 Web platform to conduct subjective listening tests with most popular evaluation standards. Optimal building of corpora using Lagrangian relaxation and a greedy algorithm. Unit selection synthesis engineĪ speech sythesis engine based on a unit selection algorithm with possibilities to propose and test new cost functions. ROOTS can be used in combination with standard tools like Transcriber and Wavesurfer, widely used in the speech community.īrowse Roots’ website to download the toolkit and to get more information. The library is coming along with scripts and wrappers enabling quick basic operations on a speech signal (ranging from linguistic to acoustic). Beside C++, a complete Perl binding has been developed. ROOTS, standing for Rich Object Oriented Transcription System, is a C++ library dedicated to annotated sequential data management, usually speech or text.
0 Comments
Leave a Reply. |