Collaborative Geospatial Data


Open glue for the web of data

Stuart Harrison ยท @pezholio

Geo data is brilliant

  • Easily reference-able
  • Great for visualisations
  • Even better for linking datasets together

HOWEVER

Many organisations

Closed datasets

Padlock by Simon Cocks (CC BY-NC-SA 2.0)

Enter OpenStreetMap

  • Not appropriate for temporary information (events etc)
  • No direct access to raw data via URIs
  • Sharealike licence
  • Centralised
Tim Berners-Lee in hi viz vest by Ivansanchez (CC BY-SA 2.0)

The solution(s)?

  • Capture existing openly licensed data
  • Extract into a common format
  • Crowdsource the gaps

Open Green Spaces

  • Taking green space data from various sources
  • Allowing the public to add extra context and locations
  • Licensed under the UK Open Government Licence

Open Addresses

  • Building upon Open Green Spaces
  • Taking existing OGL-licensed content:
    • Land Registry price paid data
    • Companies House
    • OS Locator
    • CodePoint Open

Closing the gaps

  • Programatically guessing gaps (e.g. house numbering)
  • Crowdsourcing sources of data from other organisations
  • Allowing individual edits (similar to OSM)
  • Confidence scoring: How reliable is this data point?

The future

Stuart Harrison Sam Pikesley James Smith Jeni Tennison

Open Data Institute Tech Team
@ukoditech
info@theodi.org
irc.freenode.net #theodi

ODI

http://theodi.github.io/presentations

ODI Creative Commons