The Australian Government coat of Arms

Communities of practice

Communities of practice

Data sharing between agencies


#1

In the absence of a “Closed Data” forum I’ll ask here: does anyone have experience with, or knowledge of, a data sharing platform for inter-agency sharing of sensitive data?

Data Republic provides such a platform but I suspect there might be demand for such a service from many agencies, and therefore something might already exist specifically for government use.

The platform would have to provide a mechanism by which an agency can define the policies for access to the data by other agencies.

Thanks,

Michael


#2

Treasury and ABS use Sig-box for sharing embargo data. Well established platform.

archTIS is cool if you need strong collaboration and editing.

Julian


#3

Thanks for the great info Julian!

These are both “file sharing” (i.e. unstructured data) services, whereas I was hoping for a way to share structured data so that access can be controlled at the row/column level (assuming relational data).

I think I might be asking too much :wink:


#4

Hi Michael,

We’re currently working on an Alpha version of something like this with Data61, building on the MAGDA software that’s now on data.gov.au to improve search and discovery inside/across a couple of agencies.

There’s the Data61 Protari work, but that’s more about access to data inside a secure environment.

Finance was also working on something that might fit your description, but I don’t know where it’s at currently.

Aside from the large DIPA projects like BLADE and MADIP, that’s about all I’ve come across so far.

Please get in touch with me at gordon.williamson at dta.gov.au if you’d like to discuss further.

Gordon


#5

Oh right. We use different products for that and they may not be of use to you.

We have Tablebuilder (https://abs.gov.au/websitedbs/D3310114.nsf/home/About+TableBuilder) for access to microdata, both public (open) and registered access and we also have APIs for other data sharing using .Stat (https://siscc.org/what-we-do), both open and authenticated.

We also have a secure virtual datalab, where researchers can view more granular microdata and run their own programs.

These are tailored to statistical data, including reducing the risk of disclosure, which may be specific to our use case. We have sensitive, but unclassified data.

I haven’t seen a wider, separate platform in use.