2nd Oct - Collective notes



DDI Lifecycle

Sendai Terms and Concepts

List of terms agreed by the UN in relation to Sendai.  This needs to be made into a (web-referenced) vocabulary where teach term has a URI and those which partiuclarly relate to things that are counted / measured / estimated identified as the basis for a shared vocabulary for data standardisation.

e.g. list of Causes of Mortality (from a table in a manuscript from Sendai group)  or the same set in tree view  


Need an infrastructure to support data integration, but funding for data infrastructure is hard to obtain.

What is the data model / standards in resilience.io ?

We can automate the discovery of data and this does not necessarily require a lot of metadata.

Armed with ontologies, it is possible to query across many heterogenous datasets.

Platonic view of ontologies.  Fragments of the form of ontology, they are mere representations.  But they may be useful.

Fraunhofer: semantic data lake.  The 'data lake' is checked for candidate data sets through multiple sparql queries.

Plinth Recipe: Human developed ontologies.  Machine learning / web crawling.  Proven query mediation.  Good API design.

Presupposition is the ability to link the column headings to their definition such that the query can understand this.

UKDA has started doing some work along similar lines.  Would be useful to have more information about this and what precisely it involved > Steve McEachern.

Jay Greenfield

PROV-O and DDI 4 Data Management View.  

International household survey - tools to assist creation of metadata.  This guidance and training references the terms in the DDI 4 model.