Using ZPar-MVT Off-the-Shelf

Overview

ZPar for MVT (Peking University Multi-view Chinese Treebank) is compiled into a program: zpar.mvt. This program needs to be executed with a set of corresponding statistical models. Some example sets of models are released together with ZPar source so that the public release can be used off-the-shelf.

The current version of ZPar is 0.7. Its release contains a set of models for zpar.mvt, which support joint Chinese word segmentation and POS tagging, and labeled dependency parsing.

Download and installation

The source code and models can be downloaded from sourceforge. Unzip the source zip file into the source directory and the corresponding model files each into a model directory.

Download the models for ZPar-MVT:
Download the models for ZPar-MVT.

To compile ZPar-PMT, type make zpar.mvt in the zpar source directory. The binary file zpar.mvt will placed in the dist folder.

Usage of ZPar for Chinese-MVT

Suppose that the executable files are saved in the folder zpar/dist/zpar.mvt and the models are saved in chinese.mvt. To run zpar, type 

zpar/dist/zpar.mvt chinese.mvt input output

to read Chinese sentences from the input and write the corresponding parses to output. In the file input, each line should contain only one sentence.

Annotation Schema

The annotation schema for word segmentation, POS tagging and dependency parsing is described in our COLING2014 paper (see in Reference). It is based on the annotated corpus of People' Daily of Peking University and quite different from CTB. For details, please refer to our paper.

Reference