DeSR

DeSR (c) Giuseppe Attardi 2005-2012 DeSR is a statistical dependency parser for natural languages. DeSR can be trained from annotated corpora like those supplied in the CoNLL 2006 and 2007 shared tasks. 1. INSTALLATION Issue configure from this directory: > configure then build the parser with: > make This will produce the following program: src/desr If compiling on a 64-bit machine, code that will run both on 32-bit and 64-bit machines can be generated by issuing: > configure enable-32bit before performing make. The 64-bit version is required though if you need to handle models larger than 4GB. 2. CLASSIFIERS Several classifiers are included in the package: Multi Layer Perceptron (with help from Joseph Turian) Maximum Entropy SVM (using libSVM code from http://www.csie.ntu.edu.tw/~cjlin/libsvm) Averaged Perceptron (by Massimilano Ciaramita) Passive Aggressive Perceptron 3. TRAINING and PARSING Training the parser requires an annotated corpus in the CoNNL tab format. The parsing model to use is supplied in a configuration file, which defaults to: desr.conf The parser models used for the CoNLL 2007 shared task are available in the conll07 directory. A visual tool for creating annotated corpora is available at: http://medialab.di.unipi.it/Project/QA/Parser/DgAnnotator 4. INFORMATION The home page for the project is: https://sites.google.com/site/desrparser/ The code is available on Sourceforge at: http://sourceforge.net/projects/desr Enjoy Giuseppe Attardi attardi@di.unipi.it