Scientext ANR Project

Version française

The Scientext project allows researchers, teachers and students to access a large corpus of academic writing in order to study its linguistic, discursive and sociolinguistic characteristics.

EndofProject Report: Click here to read the report (in French) (pdf)

Corpus Contents

The three essential parameters of this corpus are language (French and English), discipline (from the fields of the social sciences, life sciences and applied or engineering sciences), and genre (theses, articles, conference proceedings, student papers).

Corpus Size, Annotation Type and Distribution

The initial Scientext project has resulted in an online corpus of scientific writings of approximately 4.8 million words in French. The larger English corpus (33 million words) contains 13 million annotated words, following the inclusion of a large body of Open Access articles on Medicine and Biology). The Scientext corpus also contains a sub-corpus of 1.1 million words of academic texts in English as a Second Language. The corpus header (metadata on the annotation and corpus sources), structure (article sections), morphology, and syntax were annotated using Syntex, the parser developed by Didier Bourigault and following the existing XML and TEI guidelines.

The Research Groups and Researchers Involved in the Project

The initial Scientext project is the result of the collaboration of multidisciplinary research teams in the fields of linguistics, computational linguistics, and teaching methodology.

The LIDILEM (Laboratory of Linguistics and Didactics of Foreign and Maternal Languages) (F. Grossmann, A. Tutin, G. Antoniadis, F. Boch, C. Cavalla, M. Florez, O. Kraif, I. Novakova, M. Mroué, M.L. Nguyen, F. Rinck), project coordinator.

The LiCoRN (Corpus Linguistics) research group in Lorient (G. Williams, Ch. Millon).

The LLS (Language, Litterature, Society) in Chambéry (J. Osborne, A. Henderson, R. Barr).

Véronika Lux, with INIST, was the TEI expert on the projet.

Cécile Frérot, with the GREMUTS research group, was also involved in the project.

Achille Falaise, with the LIG, developed the project’s technical interface in collaboration with O. Kraif.

For any questions, contact agnes.tutin@u-grenoble3.fr, with [scientext] in the subject line.