Australia Cleans Up Data.Gov.Au, Loses More Than Half of Its 1,200 Datasets
BY Jessica McKenzie | Thursday, November 14 2013
Since Australia switched their open data website to an open source platform in July, the number of datasets has dropped from 1,200 to 500. Did they get lost in the move? No, it just turns out that many of them were pure junk, links to webpages that no longer existed or led to irrelevant pages.
The country launched data.gov.au in March 2011 with 200 public sector information datasets available, either hosted in the cloud or available through links to agency sites. In the two years since it grew to 1,200 data sets, or so they thought.
“We unfortunately found that a third of the “datasets” were just links to webpages or files that either didn’t exist anymore, or redirected somewhere not useful to genuine seekers of data,” the Australian Government CTO John Sheridan explained in a blog post.
Another factor in the reduction of datasets was that previously each individual file was counted as a dataset, even when two or more files were part of the same dataset.
“On the new platform,” Sheridan writes, “a dataset may have multiple files. In one case we have a dataset with 200 individual files where before it was counted as 200 datasets.”
Currently data.gov.au contains 529 datasets from 123 different government organizations.
Earlier this week, they launched a data request site where citizens can request specific government data to be released and vote for other user-submitted requests to add support.
Personal Democracy Media is grateful to the Omidyar Network and the UN Foundation for their generous support of techPresident's WeGov section.