Hello Philip, thank you for the response. Can you please provide a list of languages that can be handled by KIM out of the box or following the given process?
Regards, Naaman Musawwir. -----Original Message----- From: Philip Alexiev [mailto:philip.alex...@ontotext.com] Sent: Wednesday, February 15, 2012 5:35 PM To: Naaman Musawwir Cc: kim-discussion@ontotext.com Subject: Re: [Kim-discussion] Processing Spanish Laguange Documents Hello Naaman, KIM does not support Spanish out of the box. The knowledge base of the public distribution contains information for more generic information extraction, like world news. It contains the most famous Persons, Locations and Organizations around the world. Probably this knowledge is too general for your domain of interest. Various amount of tuning is required. There are generally two aspects of the named entity recognition process. Gazetteer Lookups This is the process of recognizing well known objects in the text. You need to supply a comprehensive set of named entities in spanish, which will feed the gazetteer. Then the gazetteer will be able to match them in the texts and create the corresponding annotations over them. More information on this can be found in KIM's system documentation* under Administration -> Extending the KIM ontology and knowledge base. New Entities This is a more complicated and composite approach, combining different techniques and rules. An example is using titles like "Mayor" and "Mrs.". You can start by looking at the grammars that are being loaded (KIM/context/default/resources/grammar/main/main.jape) and the rules that create annotations. These rules may include direct text matching of the context around the annotation or matching previously created annotations. * http://www.ontotext.com/sites/default/files/kim/KimDocs-3.0-EN.zip Hope this helps Philip Alexiev Software Engineer, KIM team On 14 Feb 2012, at 3:30 PM, Naaman Musawwir wrote: > Hello, > > We are going to try keyword extraction on some documents those are in > Spanish Language. Please direct how to configure my KIM instance to do that. > > Regards, > Naaman Musawwir. > > > _______________________________________________ > Kim-discussion mailing list > Kim-discussion@ontotext.com > http://ontomail.semdata.org/cgi-bin/mailman/listinfo/kim-discussion ----- No virus found in this message. Checked by AVG - www.avg.com Version: 2012.0.1913 / Virus Database: 2112/4810 - Release Date: 02/14/12 _______________________________________________ Kim-discussion mailing list Kim-discussion@ontotext.com http://ontomail.semdata.org/cgi-bin/mailman/listinfo/kim-discussion
