Section: New Results
Participants : Marie Candito, Benoit Crabbé, Pascal Denis, François Guérin.
Dependency trees are often preferred to syntagmatic trees for many NLP tasks, such as information extraction, question answering, lexical acquisition. We started in 2008, and continued in 2009, to work on the conversion of the syntagmatic trees of the French treebank into surface dependency trees. We have now a stabilized version of a dependency treebank : the French treebank converted to dependencies  .
The constituent-to-dependency conversion procedure can also be applied to syntagmatic trees as output by a parser trained on the syntagmatic treebank. Hence, we have various ways to obtain a parser outputting dependency trees : (i) training a parser on syntagmatic trees, and converting the output of this parser into dependencies  . And (ii) directly using existing algorithms to train a dependency parser on the treebank converted to dependencies. We have begun a comparison of the two approaches. First bare results  show for now that this second approach leads to better results : directly training a dependency parser with the MST algorithm  outperforms the architecture where a parser is trained on the French treebank (using Petrov's algorithm), and output trees from this parser are converted to dependencies. We plan to work on a more qualitative comparison of the strength and weaknesses of both approaches.