Section: New Software and Platforms
Syntax
Participants : Pierre Boullier [correspondant] , Benoît Sagot.
See also the web page http://syntax.gforge.inria.fr/ .
The (currently beta) version 6.0 of the Syntax system (freely available on Inria GForge) includes various
deterministic and non-deterministic CFG parser generators. It includes in particular an efficient implementation of
the Earley algorithm, with many original optimizations, that is used in several of Alpage's NLP tools, including the
pre-processing chain Sx Pipe and the LFG deep parser SxLfg . This implementation of the Earley algorithm has been
recently extended to handle probabilistic CFG (PCFG), by taking into account probabilities both during parsing (beam)
and after parsing (
In 2014, an in-depth rewriting of the RCG parser has started, in order for RCG parsers produced by SYNTAX to handle input DAGs while remaining efficient [60] , although parsing time complexity might, on such inputs, become exponential w.r.t. their length, whereas RCGs exactly cover the set of languages that are parsable in polynomial time (if the input is a string).
Direct NLP users of Syntax for NLP, outside Alpage, include Alexis Nasr (Marseilles) and other members of the (now closed) Sequoia ANR project, Owen Rambow and co-workers at Columbia University (New York), as well as (indirectly) all Sx Pipe and/or SxLfg users. The project-team VASY (Inria Rhône-Alpes) is one of Syntax ' user for non-NLP applications.