LMVR
Installation Guide
===========================
In order to use LMVR, you first need the Morfessor Baseline.
Both are installed using setuptools library for Python.
For more information, see:
http://morfessor.readthedocs.io/en/latest/installation.html
You can install them separately or use the methods from the Python project.
Download and unzip lmvr.
To build and install the module and scripts to default paths, type
python setup.py install
Running the software
--------------------
There are some example scripts in the folder lmvr/examples.
You can train a morphology lexicon and segment your corpus based on this lexicon using on the two scripts.
The scripts are generic for MT experiments, so defines as input train, dev, test sets and includes pre-processing of all corpora.
Documentation
-------------
Details of the method are described in the paper
Ataman, D., Negri, M., Turchi, M., Federico, M. (2017) Linguistically Motivated Vocabulary Reduction for Neural Machine Translation from Turkish to English. The Prague Bulletin of Mathematical Linguistics 108, 331--342, Prague, The Czech Republic, June 2017, European Association for Machine Translation.
Contact
-------
For further questions contact ataman@fbk.eu.