Metadata Evaluation and Improvement

Community Inventory of EarthCube Resources for Geoscience Interopoerability

(CINERGI) cinergi.cloudapp.net

Ilya Zaslavsky and Stephen Richard

Note: Part of this presentation contained live demonstrations, which are not attached.

• Lots of datasets and catalogues used by Earth Scientists, but there are lots of differences between them. People want to go to one site and get all of the relevant information. 

• This is a tool for metadata aggregation with:

  • Searchable metadata

  • Automated metadata enhancement (Spatial Extent, Keywords, Abstracts, etc.)

• The Enhancer for metadata contains several components

  • GeoSciGraph API is used for semantic processing

  • Allows for manual editing and review of keywords and location assignments for machine learning

• Currently working to engage community

  • Looking to get lists of resources and catalogues from all domains and communities

• CINERGI is available on GitHub

• Lots of group discussion on semantic conflicts and linking a term to multiple meanings across and within ontologies

Big Earth Data Initiative (BEDI) Metadata Improvement: Case Studies

John Kozimor

• The goal: Optimize the collection, manage and deliver of US Govenrment data to improve discovery, access use, and understanding
  • Metadata helps to achieve this
• Tool swere developed to evaluate metadata
  • Applied to NASA and USGS collections
  • 3 Case Studies
• Case Study 1
  • What metadata are important across communities?
    • Identify concepts that are common and not common across mandatory level recommendations 
    • Actually few are common
• Case Study 2
  • Do my dialects support my requirements?
    • Identify gaps between organizational capabilities (dialects and new recommendations (requirements)
      • (CSDGM, DataCite, CSW, ISO, DIF, ECHO)
• Case Study 3
  • How complete is my metadata with respect to recommendations?
    • Evaluate and measure the completeness of NASA and USGS metadata collections with respect to community recommendations
    • This sparked a decent amount of discussion over what a “required recommendation” means, and if all dialects can comply.

Metadata Evaluation & Improvement, Case Studies

            Ted Habermann

• Discussing UMM

  • Metadata Quality: Completeness and Consistency

• Looked through a lot of metadata

  • Asked specific questions

• Comparisons between various datasets and what metadata is contained

  • Things like:

    • Role, Point of Contact, Distribution Contact, Processor, Email Address, etc.

  • Lots of discussion on what Email Addresses should be included.

• Next Step – Metadata Components

  • Decided on important content

  • Create consistent content for people and organizations and connect to CMR

•