BENGUELACC OIH Meeting 2022 04 28

Meeting time

• 2022-04-28 @ 12:00 UTC
• 60 minutes in length

Meeting Location

• Jitsi meeting (audio or optional video)
  • connect directly in your browser (no installation required) by clicking: https://meet.jit.si/OceanInfoHub
  • or use the "Jitsi Meet" app on Android or iOS
    • room name: OceanInfoHub

Meeting Agenda

• Agenda1: introductions
• Agenda2: background on ODIS
  • https://book.oceaninfohub.org/
• Agenda3: background on BCC Geodata Portal
  • https://geodata.benguelacc.org
• Agenda4: potential steps for connecting BCC with ODIS
• Agenda5: schedule next meeting

Action Items

Action Item	Assigned to
Review ODIS thematic patterns & provide feedback	BCC team
Possibly run the Python JSON-LD process locally	BCC team
Adapt thematic patterns (or create new) accordingly	Jeff & Doug
Adapt Python JSON-LD process to BCC feedback	Jeff

Log

• Attendance: Anja (BCC), Zukile (BCC), Mika (ODINAfrica), John (ODINAfrica), Pier-Luigi (OIH), Doug (OIH), Jeff (OIH)
• agenda: background
  • Anja: admin who setup the GeoNode portal: https://geodata.benguelacc.org/
  • Zukile:
    • want system to accommodate many sectors
    • want data to be useful
    • some private data
    • end-of-June publish the portal/system
  • PL:
    • OIH is a way to harvest
    • schema.org based
    • contains 6 main patterns: https://book.oceaninfohub.org/thematics/README.html#thematic-profiles
      • Experts and Institutions
      • Documents
      • Spatial Maps
      • Projects
      • Training
      • Vessels
  • Doug:
    • discussing the ODIS 3-step process the expose your catalogue layers through schema.org, index, and then discovery
      • 
  • Jeff:
    • exposing your catalogue layers through JSON-LD gives 2 big benefits:
      • that record/dataset will be discoverable through the upcoming "Google Dataset Search" tool: https://datasetsearch.research.google.com/
      • your layers will be shared through ODIS and discoverable & queryable by other nodes (ocean data centres) around the world
  • Zukile:
    • discoverability is important
• agenda: potential steps for connecting BCC with ODIS
  • Jeff:
    • GeoNode does not yet automatically include JSON-LD for the layers
    • a workaround is to generate the JSON-LD for each layer in your catalogue
      • Python3 script: https://github.com/iodepo/odis-arch/blob/schema-dev-jm/code/notebooks/Exploration/benguelacc-harvest.py
      • see sample JSON-LD output for your GeoNode layers: https://github.com/iodepo/odis-arch/tree/schema-dev-jm/code/notebooks/Exploration/data-benguelacc
    • you would run that script, and then add each JSON-LD into your sitemap.xml file
      • sample sitemap.xml of your layers: https://github.com/iodepo/odis-arch/blob/schema-dev-jm/code/notebooks/Exploration/data-benguelacc/sitemap.xml
      • there is also a script to generate that sitemap.xml file, in case you find it useful: https://github.com/iodepo/odis-arch/blob/schema-dev-jm/code/notebooks/Exploration/generate_sitemap.php
      • you would then submit that sitemap through your Google Search Console: https://search.google.com
  • Anja:
    • We went with GeoNode as a catalogue server, but is GeoNetwork better? (we have a lot of time-series data etc, that is not necessarily geospatial)
  • Jeff:
    • stick with GeoNode (it takes time for some changes to trickle down, but JSON-LD is coming through pycsw likely)
    • [note added later] : CKAN is a very popular all-around Open Source catalogue server, very popular in Canada at federal government level, but it is not necessarily geospatial-based / there are extensions for spatial : https://ckan.org/
      • the Pacific region groups as part of ODIS leverage CKAN : see https://pacificdata.org/
      • but honestly, your GeoNode system is great, switching would require a lot of effort (and the same 'we need JSON-LD native support' discussions are happening now in CKAN)
  • Anja:
    • we have a lot of species data, a lot of metadata....is there a template for metadata we can follow?
  • PL:
    • the templates are the 6 patterns I mentioned, examine those for your data & let us know if you require additional 'patterns'
• schedule next meeting
  • Zukile:
    • May is a very busy month, we need time to get our house in order
  • next meeting date set: 2 June