There are long-established connections between research on the sound shape of language and natural language processing (NLP), for which one of the main driving forces has been the design of automatic speech synthesis and recognition systems. Over the last few years, these connections have been made yet stronger, under the influence of several factors. A first line of convergence relates to the shared collection and exploitation of the considerable resources that are now available to us in the domain of spoken language. These resources have come to play a major role both for phonologists and phoneticians, who endeavor to subject their theoretical hypotheses to empirical tests using large speech corpora, and for NLP specialists, whose interest in spoken language is increasing. While these resources were first based on audio recordings of read speech, they have been progressively extended to bi- or multimodal data and to spontaneous speech in conversational interaction. Such changes are raising theoretical and methodological issues that both phonologists/phoneticians and NLP specialists have begun to address.
Research on spoken language has thus led to the generalized utilization of a large set of tools and methods for automatic data processing and analysis: grapheme-to-phoneme converters, text-to-speech aligners, automatic segmentation of the speech signal into units of various sizes (from acoustic events to conversational turns), morpho-syntactic tagging, etc. Large-scale corpus studies in phonology and phonetics make an ever increasing use of tools that were originally developed by NLP researchers, and which range from electronic dictionaries to full-fledged automatic speech recognition systems. NLP researchers and phonologists/phoneticians also have jointly contributed to developing multi-level speech annotation systems from articulatory/acoustic events to the pragmatic level via prosody and syntax.
In this scientific context, which very much fosters the establishment of cross-disciplinary bridges around spoken language, the knowledge and resources accumulated by phonologists and phoneticians are now being put to use by NLP researchers, whether this is to build up lexical databases from speech corpora, to develop automatic speech recognition systems able to deal with regional variations in the sound pattern of a language, or to design talking-face synthesis systems in man-machine communication.
The goal of this special issue will be to offer an overview of the interfaces that are being developed between phonology, phonetics, and NLP. Contributions are therefore invited on the following topics:
Joint contributions of speech databases to NLP and phonology/phonetics
Automatic procedures for the large-scale processing of multi-modal databases
Multi-level annotation systems
Research in phonology/phonetics and speech and language technologies: synthesis, automatic recognition
NLP and modelisation in phonology/phonetics
Papers may be submitted in English (for non native speakers of French only) or French and will relate to studies conducted on French, English, or other languages.
11 February 2008: Extended deadline for reception of contributions
11 April 2008: Notification of pre-selection / rejection
11 May 2008: Reception of pre-selected articles
16 June 2008: Notification of final acceptance
30 June 2008: Reception of accepted articles’ final versions
This special issue of Traitement Automatique des Langues will appear in autumn 2008.
Contributions (25 pages maximum, PDF format) must be sent by e-mail to the address below:
Style sheets are available for download on the Web site of the journal.
TAL (Traitement Automatique des Langues / Natural Language Processing) is a forty-year old international journal published by ATALA (French Association for Natural Language Processing) with the support of CNRS (French National Center for Scientific Research). It has moved to an electronic mode of publication, with printing on demand. This affects in no way its reviewing and selection process.
Martine Adda-Decker, LIMSI, Orsay Roxane Bertrand, LPL, CNRS & Université de Provence Philippe Blache, LPL, CNRS & Université de Provence Cédric Gendrot, LPP, CNRS & Université de Paris III John Goldsmith, University of Chicago Guillaume Gravier, Irisa, CNRS/INRIA & Université de Rennes I Jonathan Harrington, IPS, University of Munich Bernard Laks, MoDyCo, CNRS & Université de Paris X Lori Lamel, LIMSI, Orsay Noël Nguyen, LPL, CNRS & Université de Provence François Pellegrino, DDL, CNRS & Université de Lyon II François Poiré, University of Western Ontario Yvan Rose, Memorial University of Newfoundland Tobias Scheer, BCL, CNRS & Université de Nice Atanas Tchobanov, MoDyCo, CNRS & Université de Paris X Jacqueline Vaissière, LPP, CNRS & Université de Paris III Nathalie Vallée, DPC-GIPSA, CNRS & Université de Grenoble III