Summary of the paper
| Title | Statistical French Dependency Parsing: Treebank Conversion and First Results |
|---|---|
| Authors | Marie Candito, Benoît Crabbé and Pascal Denis |
| Abstract | We first describe the automatic conversion of the French Treebank (Abeilléand Barrier, 2004), a constituency treebank, into typed projective dependencytrees. In order to evaluate the overall quality of the resulting dependencytreebank, and to quantify the cases where the projectivity constraint leads towrong dependencies, we compare a subset of the converted treebank to manuallyvalidated dependency trees. We then compare the performance of twotreebank-trained parsers that output typed dependency parses. The first parseris the MST parser (Mcdonald et al., 2006), which we directly train ondependency trees. The second parser is a combination of the Berkeley parser(Petrov et al., 2006) and a functional role labeler: trained on the originalconstituency treebank, the Berkeley parser first outputs constituency trees,which are then labeled with functional roles, and then converted intodependency trees. We found that used in combination with a high-accuracy French POS tagger, the MST parser performs a little better forunlabeled dependencies (UAS=90.3% versus 89.6%), and better for labeleddependencies (LAS=87.6% versus 85.6%). |
| Language | Statistical and machine learning methods |
| Topics | Corpus (creation, annotation, etc.), Parsing, Statistical and machine learning methods |
Full paper  |
Statistical French Dependency Parsing: Treebank Conversion and First Results |
| Bibtex | @InProceedings{CANDITO10.392, author = {Marie Candito, Benoît Crabbé and Pascal Denis}, title = {Statistical French Dependency Parsing: Treebank Conversion and First Results}, booktitle = {Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10)}, year = {2010}, month = {may}, date = {19-21}, address = {Valletta, Malta}, editor = {Nicoletta Calzolari (Conference Chair), Khalid Choukri, Bente Maegaard, Joseph Mariani, Jan Odjik, Stelios Piperidis, Mike Rosner, Daniel Tapias}, publisher = {European Language Resources Association (ELRA)}, isbn = {2-9517408-6-7}, language = {english} } |