Hi Tomek, I would advise you to involve the OpenNLP community, as they have already done a lot of work with NER for other languages: http://opennlp.sourceforge.net/models-1.5/
A good starting point would be to train OpenNLP models for Polish based on Wikipedia data, as Olivier once shared here: http://dev.blogs.nuxeo.com/2011/01/mining-wikipedia-with-hadoop-and-pig-for-natural-language-processing.html Cheers, Pablo On Tue, Sep 25, 2012 at 3:16 PM, Fabian Christ <[email protected] > wrote: > Hi Tomek, > > and welcome! Nice to hear that you have to offer an engine and would > like to contribute. > > To contribute your engine, you should open an issue in the JIRA system > [1] describing your contribution, how to use it, how to install it. > Then you should upload a patch file to the issue that contains your > changes (the source code). A Stanbol committer will take this up and > try to apply the patch. In this way your contribution finds its way > into the Stanbol code base. > > Best, > - Fabian > > [1] https://issues.apache.org/jira/browse/STANBOL > > 2012/9/25 Tomek <[email protected]>: > > Hi! > > My name is Tomek and I am student from Poznań University of Technology > in Poland. > > Recently I worked for PSNC. My task was to create NER engine for polish > language. > > My supervisor asked me to contact with you and include my work in > stanbol project. > > Could I get some information how to add new feature to your project (if > You are interested). > > > > Tomek > > > > -- > Fabian > http://twitter.com/fctwitt > -- --- Pablo N. Mendes http://pablomendes.com Events: http://wole2012.eurecom.fr
