JEP-TALN-RECITAL 2016, Paris, France

23rd French Conference on Natural Language Processing
31st "Journées d’Études sur la Parole"
18th Meeting of Student Researchers in Computer Science for Natural Language Processing

Inalco, Paris, 4-8 Juillet 2016


The Greater Paris Area laboratories working on speech and natural language processing (written, spoken and signed languages) are jointly organizing the fifth joint JEP-TALN-RECITAL conference at INALCO (inner Paris) on

July 4-8, 2016

For the fifth time, after Nancy (2002), Fez (2004), Avignon (2008), Grenoble (2012), AFCP (the French Association for Spoken Communication) and ATALA (the French Association for Natural Language Processing) are jointly organizing their main conferences to gather in one venue the research communities of written, spoken and signed language processing.

This edition gathers:

  • the 31st Speech Research Days (Journées d'Etudes sur la Parole - JEP),
  • the 23rd French Conference on Natural Language Processing (Traitement Automatique des Langues Naturelles - TALN),
  • the 18th Student Research Meeting on Natural Language Processing (Rencontre des Étudiants Chercheurs en Informatique pour le Traitement Automatique des Langues - RÉCITAL).

TALN Topics

The submitted papers must address topics related to NLP, including but not limited to:

  • Phonetics, phonology, morphology, part-of-speech tagging
  • Syntax, grammars, parsing, chunking,
  • Semantics, Pragmatics, Discourse,
  • Lexical and distributional semantics,
  • Linguistic and psycho-linguistic aspects of NLP,
  • Resources for NLP,
  • Evaluation methods for NLP,
  • NLP applications (information retrieval and extraction, question-answering, generation and summarization, dialogue, sentiment analysis, etc.),
  • NLP and multi-modal interface (speech, vision, etc.),
  • NLP and multilinguality,
  • NLP for the Web and social networks,
  • NLP and low-resourced language,
  • NLP and signed language,
  • NLP and ethics.

Types of submissions

Two formats of papers are proposed: long papers (12 pages plus up to 2 pages for references) and short papers (6 pages plus up to 2 pages for references). Long papers will be presented orally and short papers as posters.

TALN solicits two types of communications:

  • papers presenting original research work
  • position statements presenting a point of view on the state of NLP research

Submissions in English are acceptable if one of the authors is not a French speaker. The accepted papers will be published in the conference proceedings.

Submissions must present original, unpublished research. Thus, the translation of an already published paper is not acceptable.

Selection criteria

Each submission will be reviewed by at least three specialists of the domain. The following criteria apply:

  • appropriateness to the conference topics,
  • substance and originality of the contribution,
  • correctness of the technical and scientific content,
  • critical discussion of the results, in particular with respect to the state of the art,
  • position of the work in the context of international research,
  • organization and clarity of presentation.

For position statements, the following qualities are emphasized:

  • scope of the presentation and consideration of the state of the art
  • originality and impact of the presented point of view.

A committee composed of members of the TAL Journal Editorial Board and of the TALN Program Committee will select, among the accepted papers, one paper (TALN Award) which will be recommended for publication (as an extended paper) in the journal Traitement Automatique des Langues (TAL).

Submission procedure

Submitted papers must conform to the official guidelines and style as described in the style file (https://jep-taln2016.limsi.fr/styles/jeptaln2016-v2.zip) and must be in PDF format.

Submission web site for the demonstration papers: https://easychair.org/conferences/?conf=jep2016 (track JEP-TALN 2016 System Demonstration)

Submission web site: https://easychair.org/conferences/?conf=taln2016

Since the review process is double-blind, the submitted papers must not include the authors' names and affiliations. Furthermore, self-references and project names that reveal the author's identity, e.g., "We previously showed (Smith, 1991) ...", must be avoided. The authors should use instead citations such as "Smith (1991) previously showed ...". Acknowledgments should not be included in the first submission, and can be added instead to the final version of a paper after acceptance.