Hi Olivier, at the moment the software is on our own svn, its commented and there is some short documentation. Alberto can give you more details, and if you want also make you access to the svn. Please, liaise with him directly, it's faster :)
I think the best is to look at, and discuss about it at the hackathon and possibly decide how/if to integrate it in Stanbol. The components are already provided as REST services, but its integration is not straightforward. The main difference I understand between your approach and ours is that we exploit identity resolution results for the categorization and we create the index by selecting certain parts of the texts. The index is generated through SPARQL queries, hence it can be customized. Said that maybe your component can be extended with parts of ours if we realize that it can be improved. Valentina On Nov 18, 2011, at 4:53 PM, Olivier Grisel wrote: > 2011/11/18 valentina presutti <[email protected]>: >>>> On Tue, Nov 15, 2011 at 12:45 PM, Stefane Fermigier <[email protected]> wrote: >>>>> Is online here: >>>>> >>>>> >>>> http://www.slideshare.net/nuxeo/apache-stanbol-and-the-web-of-data-apachecon-2011 >> >> Very nice presentation Olivier :) >> please have a look at [1], it's a demo we showed some time ago. >> It also makes some general classification based on Wikipedia categories. >> I think there's some procedure we use for creating the index that can be >> integrated in this component, it could possibly further improve the results. > > Interesting. Is there a technical description or the source code > available online? > > -- > Olivier > http://twitter.com/ogrisel - http://github.com/ogrisel
