PoeM: Provenance Mining for scientific linked data publishing
Contact : Alban Gaignard, Hala Skaf, Audrey Bihouee, {firstname.lastname}@univ-nantes.fr.
For demonstration purpose only.
This web page illustrates our semi-automated for mining provenance traces and assembling linked experiment reports. PoeM generates Semantic Web rules from (i) annotated workflow patterns, (ii) domain-specific annotations, and (iii) provenance traces of a workflow run. The rules finally match provenance subgraphs and produce linked experiment reports.
We illustrate our approach in the context of an RNAseq bioinformatics .
This work reuses existing linked open vocabularies, namely PROV-O, P-PLAN, Micropublications Ontology, Experimental Factors Ontology, and EDAM.
This demo is supported by the Corese Semantic Web factory, Apache Jena, D3.js, Codemirror.js and Twitter Bootstrap.