Text Mining Task Force

Members

•  Alberto Labarga
•  Hammad Afzal
•  Matthias Samwald
•  Venkata Satagopam
•  Heiko Horn
•  Hong-Woo Chun
• Takagai-san
• Ohuchida-san

Proposal

• Integration of text mining results
  •  Reflect /  Whatizit /  DBCLS NER
  • Results as triplets  Check Science Commons text annotation service
    • URI and predicates
  • Implementation with  SADI
  • Result presentation using  <a>Tag
• Explore relations
• Interfaces

Ideas

Reflect RDFa snippets

Idea 1 (interlaced with the text):

<div xmlns:sioc="http://rdfs.org/sioc/ns#" typeof="sioc:Item">
<span class='reflect_protein'><a rel="sioc:topic" resource="http://purl.uniprot.org/uniprot/P05113"></a>IL-5</span>
</div>

Idea 2 (at the end of the document, non-redundant):

<div xmlns:sioc="http://rdfs.org/sioc/ns#" typeof="sioc:Item">
<span id='reflectEntity0'>9606.ENSP00000231454</span>
<a rel="sioc:topic" resource="http://purl.uniprot.org/uniprot/P05113"></a>
</div>

Whatizit RDFa service

IL-5 Example:  http://whatizit.neurocommons.org/index.py/text_query?pipeline=whatizitEBIMedDiseaseChemicals&query=IL-5&format=atag&max_results=10

Converted to Turtle:  http://www.w3.org/2007/08/pyRdfa/extract?uri=http://whatizit.neurocommons.org/index.py/text_query?pipeline=whatizitEBIMedDiseaseChemicals&query=IL-5&format=atag&max_results=10

This uses a W3 service ( http://www.w3.org/2007/08/pyRdfa/extract)