ELRA, the European Language Resources Association, has announced the availability of a new spoken language resource: S0156 ANITA (Audio eNhancement In Telecom Applications), consisting of 41 recordings (17 males and 24 females) stored on 13 CDs. It comprises voice recordings in 4 languages (English, French, German and Spanish), noise recordings (sirens, engines, roadworks, crowds, trains, etc.), and real condition recordings (voices and mixed noises), in English. Each language consists of 60 phonetically rich sentences (normal and stress and in panic conditions), letters and numerals (normal and stress and in panic conditions) and a 10 minute text (normal conditions).
The language resources available in ELRA’s catalogue are distributed into 4 categories : ‘Speech and Related Resources’, ‘Written Resources’, ‘Terminological Resources’, and ‘Multimodal/multimedia resources’.
1/Spoken LRs
A – Telephone or Desktop/Microphone recordings
The databases catalogued in this section have been produced with speakers’ recordings made over the telephone (fixed or mobile) network, or through a microphone.
You will find speech resources recorded in various environments, and covering a large number of European and non-European languages, e.g. the databases produced in the framework of the SpeechDat project, the BABEL databases, produced in the framework of the Copernicus programme, consisting of spoken corpora recorded in a controlled office environment, etc.
B – Speech Related Resources
You will find in this section e.g. pronunciation and phonetic lexicons, the BDLEX, PHONOLEX, and MHATLEX databases, etc.
2/Written LRs
A – Corpora
This section contains monolingual and multilingual corpora, parallel or not, which may also be annotated. A few examples of the kind of resources you will find in this sectiion are e.g. the corpora developed in the framework of the MULTEXT project, the Multilingual and Parallel Corpora (MLCC), French scientific corpora, newspaper corpora in Arabic, etc.
B – Monolingual lexicons
The section dedicated to monolingual lexicons contains various types of dictionaries, e.g. a dictionary of French verbs, the Japanese word dictionary, some PAROLE lexicons in many languages, etc.
C – Multilingual lexicons
Here you can find either bilingual or multilingual dictionaries and lexicons, including e.g. the EuroWordNet database, etc.
3/Terminological LRs
Monolingual, bilingual and multilingual terminological databases are available. They cover a large number of specialised domains, e.g. automobile engineering, insurance, linguistics, finance, etc., in a wide variety of languages.
4/Multimodal/Multimedia LRs
The resources you will find in this section have been produced using different modalities, including the speech. An example of such resources is the database produced in the framework of the M2VTS project.
2004-02-13
Notícias