Metadata Hack-A-Thon

Metadata Hack-a-thon

Purpose is to collaborate & discuss – can break into groups

https://docs.google.com/document/d/1FBT8RGUbDFQHYPtjD9DYGcN_iMgPd5aoZ1oxITv-kpQ/edit

1)      DQ in 19157 - 6

2)      Relationship between provenance and ISO

3)      Utilization of metadata to facilitate DOI citations – 5

4)      HTML displays of metadata leveraging schema.org tags (crosswalking tags to ISO metadata) - 14

5)      xlinks in metadata to linked open data terms - 6

6)      generation of metadata from domain specific databases: questions, approaches? – 2

7)      web applications for exploring metadata (hyperlink trees, cross referencing, discovery)

8)      web services that provide metadata

9)      systems and protocols for exchanging metadata

10)  geographic representation of metadata, drilldown capability

Schema.org (#4)

·         https://schema.org/Dataset

·         how do you add a temporal tag to your dataset – tag your html that says this is a schema.org then add attributes to it

·         trying to look at “dataset”…

o   properties are : catalog, distribution, spatial, and temporal

·         example by Stephan Zednik using opendap with schema

·         use is to make your data more discoverable through google

·         maybe follow up on the semantic cluster

xlinks in metadata to linked open data terms (#5)

·         can change to anchor tag

·         the content can still be human readable

·         what if something like data.gov used these anchor

·         example has individual where position is listed

·         care should be taken in how to make it (Christine White)

·         for peoples – this would be a good way to disambiguate names – like an orchid ID

o   currently use component and reuse these

·         Cindy - make xlinks between potentially related resources and uuid

o   Any tools that can use iso metadata – would be better if iso used unique identifier

DOI citations (#3)

·         At NOAA, every time we mint a DOI there has to be a robust record associated with it.  Potential can use the metadata to automatically mint the doi – 4 tests –

o   Creator, pub year, title, publishers

o   Then have a landing page to help describe the dataset more