UIMA Semantic annotators for Nuxeo

UIMA Semantic annotators for Nuxeo

Nuxeo is working on various ways to leverage the knowledge hidden in the content with semantic technologies. The following blog post gives a global overview of what's cooking in the labs:

http://blogs.nuxeo.com/dev/2009/11/iks-semantic-search-workshop-wrapup.html

Deploying the demo from a prebuilt tech preview plugin

To be done (soon).

Building the components from source

Fetch the source of the following projects from the mercurial repos:

hg clone https://hg.nuxeo.org/sandbox/scribo/
hg clone https://hg.nuxeo.org/sandbox/scribo-data/
hg clone http://bitbucket.org/fmancinelli/jenabean/
hg clone https://hg.nuxeo.org/sandbox/nuxeo-uima/

Then follow the instuctions from the README.txt file of each project when available. The generic approach is to

mvn install -Dmaven.test.skip=true

The Scribo knowledge model

The default UIMA annotators packaged for Nuxeo are based on the work done for the http://scribo.ws project. The base annotators focus on named entities extraction. Here is a summary diagram that show the main RDF / OWL structure extracted from the content and serialized in the Knowledge Base (Jena graph).