The datasets released here have been designed as part of the SemDis 2014 workshop on distributional semantics at the TALN conference and for the paper by Tanguy et al. (2015) published in the special issue of the TAL journal.

• Datasets for the lexical substitution task
• Dataset for the evaluation of distributional models on the TALN corpus, as described in (Tanguy et al., 2015)