ELRA - ELRA-U-S 0015 : OTG Corpus

You are here » Universal Catalogue » Spoken Resources » Desktop/microphone

Language Resources

Search Catalogue

Send us information

Would you like to collaborate ?
Contact Us

Languages

Catalog Reference : ELRA-U-S 0015

OTG Corpus

The OTG corpus contains recordings of real conversations between one or several tourists and a receptionist in a French tourist office (the Tourist Office of Grenoble). Conversations were recorded with a semi-clandestine procedure: tourists were not aware of being recorded while receptionists were (but were given no instruction).
They were also recorded separately by two hidden unidirectional microphones, and stored on two separate tracks by a digital recorder (DAT), resulting in two files for each dialogues.

Language: French (native)
Number of speakers: 315 tourists + 5 receptionists (H-H real dialogues)
Field: tourist information
Recording duration: 2h
Number of dialogues: 315
Number of words: 26000
Format of the audio files: WAV, sampling frequency: 16 KHz
Type of transcription: orthographic (with Transcriber)
Transcription convention: DELIC
Format of the transcriptions: XML, ASCII, PDF

The OTG corpus was produced in 2002 by the OuRAL Project ('OUtils et Ressources pour l'Analyse de la Langue') which aimed at building resources and tools for French written and spoken language processing.
It is also part of the first delivery of Parole Publique (Public Speech), a project aiming at the creation of a large corpus (orthographic transcription and morpho-syntactic annotation) of spoken French dialogues. The Parole Publique corpus is primarily intended for researches on man-machine communication.

Identification

Period of coverage :

Version : 1.0
Version history :

Production

Project : OuRAL project

Creation date : 2002

Applications

	Applications possible : Speech synthesis#Talking head synthesis#Humanoid agent synthesis
application Area : Research

Technical Informations

Fileformat : wav

Contents

Click on the arrow to display content.

speech corpus
Language(s) : French (France)
Duration : 2 hours
Source Channel : Microphone
Speech Acquisition Mode : Acoustic
Recording Environment : tourist office
Transcription Entries : Orthographic
Transcription Verification : yes