Section: Overall Objectives

Highlights of the Year

• A statistical parsing architecture for Italian using MElt in a pre-processing step has obtained the best results in the EVALITA shared task on Italian parsing [35] (cf.  5.7 ).

• Two different instances of Alpage parsing architectures were ranked 2nd and 3rd at the SANCL shared task on parsing user-generated content, organized by Google [38] (cf.  5.7 and 6.4 ).

• Release of two freely available out-of-domain treebanks for French: the SequoiaBank focusing on well-edited texts such as Wikipedia, Europarl, ...; the French Social Media Bank, focusing on noisy user-generated content (Facebook, Twitter, ...), the latter being the first available dataset for Facebook in any language – cf  6.4 .