Principles

From Open NZ Wiki

Jump to: navigation, search

Assumptions

  • It is in the public interest to maximise data reuse
  • Government organisations are helpful, but we are starting a dialogue on improving automated access
  • 80% of public value will lie in 20% of data (but which data?)
  • The data should be available to everyone on the Internet, not just people in NZ


Principles

Principle Rationale Implications
New Zealand Open Data Principles
Equivalent cost: Same cost to all requesters A distributor of govt data that is more successful, should not be penalised for their success
  • All distributors need to have equivalent workload for govt provider
Minimal cost: All parties should seek to minimise cost to the taxpayer Potentially a value-adder of govt data could design a system, that requests fresh data every transaction; rather than caching.
  • Agencies will have to specify minimum "refresh rates". If an agency specifies real-time accuracy, (ie no caching) then agency will have to bear cost of every transaction
Authoritative source: The authoritative source of the data should be acknowledged The trust, integrity, obligations of the base data should be able to be verified by users.
  • Mechanism to handle this
Easy Use: Put information into easy to download, easy to manipulate files Take government data out of PDFs and other formats that are difficult to analyze with database and spreadsheet software, thereby empowering independent people and organizations to do their own research and analysis.
High Value: Priority should be given to providing better access sto high value datasets Agencies have limited funding. If New Zealanders can agree on what is greatest value, then increased focus can be put on getting those datasets online.
  • Need agreement on what is value / potential value.
Machine Readable: Datasets must be made available in machine readable, open standard formats.

Appropriately marked up HTML pages, designed for people first and machines second, make information more accessible.

Datasets that are machine readable enable automated processing. It encourages the building of downstream applications.

  • Government web pages are not brochure ware, they are well-assembled data structures for information exchange.
  • Need agreement on what are the preferred open standard formats. Suggestions (CSV/TXT, XML, KML/KMZ)
Permanent URL: Datasets must have a permanent URL. Changing the URL of a dataset increases the disruption for both humans and machines.
  • Need a Government PURL service, that allocates PURLs independent of agency domain name.
Freely Licensed: Datasets must be available under a "free to reuse" licence Ensures no one private entity can control the rights to or access to the raw, public dataset.
  • Need a policy on public dataset licensing
Personal tools