Text Mining Contest 2016 – DEFT2016

DEFT2016 is held in conjunction with JEP-TALN-RECITAL 2016. 

Logo TALN 2016

DEFT2016 will be held on 4th July 2016 – 9.00am, at INALCO, room 4.10, 4th floor.

» See the results and the proceedings


The DEFT (défi fouille de textes) workshop is a text mining competition organized since 2005 and that focuses each year on different research themes.

This 12th edition addresses the issue of indexing scientific documents in French.

Given a bibliographic record (title and abstract), the task is to produce a set of keyphrases that capture the main topics discussed in the record. This task mimics the manual indexation performed by professional indexers, which first relies on thesaurii, and possibly supplement the list of keyphrases selected from the thesaurii by keyphrases extracted from the content of the bibliographic record. The provided dataset will include bibliographic records from four different research areas (linguistics, information science, archeology and chemistry). The reference keyphrases were obtained as part of the project TermiTH ANR CONTINT 2012.


Logo LINALogo INISTLogo CNRS

Logo projet TermiTH   Logo Université de Nantes