since 20-02-2015
- Marten Postma ([email protected])
- Emiel van Miltenburg ([email protected])
the SoNaR corpus is a large Dutch corpus (http://tst-centrale.org/producten/corpora/sonar-corpus/6-85) , of which a part has been annotated with Cornetto senses (http://www2.let.vu.nl/oz/cltl/cornetto/) in the DutchSemcor project (http://www2.let.vu.nl/oz/cltl/dutchsemcor/). the goal of this project is to:
- COMPLETED:convert each file from Folia xml (http://proycon.github.io/folia/) to NAF (http://www.newsreader-project.eu/files/2013/01/techreport.pdf)
- COMPLETED:include the dutchsemcor annotations
- COMPLETED:add the Open Source Dutch Wordnet annotations
- TODO:run Dutch pipeline to add NER, NEL, newest alpino, SRL, timex and many more..
There are two main purposes of this github:
- convert a folia xml file to NAF containing wf and term layer. cd to the scripts folder and call python FoliaToNaf.py -h for information on how to use it.
- convert DutchSemcor to NAF. cd to the scripts folder and call python main.py -h for more information on how to use it.
Contents of this github:
- folder 'scripts': contains python scripts to perform conversion
- folder 'resources': contains 'base_naf.xml' which is used for the NAF conversion and 'cdb_syn_FILT.xml.lu-map', which is a mapping from Cornetto to ODWN1.0. It also contains the allwords xml files and its annotations. The folder allwords_NAF contains the processed all words files with annotations in naf.
- folder 'dutch_pipeline': contains scripts to run naf file through dutch pipeline. only created for use on our personal server.
- the all words part of dutchsemcor has been processed, but can not be distributed due to the license of SoNaR. Please contact the authors of this github for more information.
TODO list includes:
- run full conversion to naf with pipeline (to be set up)
All python code has been documented with the epydoc package (http://epydoc.sourceforge.net/) open script/html/index.html to inspect the documentation.