How Koordinates Solves the Problem of XML Metadata

Koordinates supports connected metadata sources, making it easier to publish and update XML metadata.


Posted by Hamish Campbell
September 10th, 2019

Metadata New Admin

This is part of a regular series of posts on lesser-known features of the Koordinates platform.

As any user of open geospatial data knows, metadata provision is often — to put it charitably — patchy. Many data publishers don’t provide metadata at all; those that do often provide non-standard metadata generated by their software provider.

Metadata publication, then, is a Problem. For professional data users, metadata is critical for determining whether a specific dataset is authoritative and fit-for-purpose. Metadata needs to make it clear who produced the data; where it came from; how relevant it is to a particular project; how current it is; and much more. Crucially, it needs to be both human and machine readable.  

So why is metadata provision so patchy? The main reason is that metadata is generally stored in a different database than the data itself. Also, most data portals don’t make it easy to import and associate data and metadata from two different data sources. Finally, data portals don't always support multiple metadata standards (such as ISO /19915/19139, GDC CSDGM, and Dublin Core, the formats supported at Koordinates).

These problems are compounded for data that is regularly updated. Bad metadata management makes monthly updates a pain, and weekly updates near-impossible. From a data manager's point of view, then, publishing XML metadata is more complicated than it seems. 

At Koordinates, we sought to address this Problem as part of a recent refresh of Data Management. Our aim was to make it easier for data managers to import and update XML metadata. 

The first improvement we made was to support XML metadata sources. This meant that data publishers could connect and import their metadata using the same workflows as for geospatial data.  (If you want to learn more about connected data sources, check out our blog from a few weeks back.)  Data managers can also now import metadata from a dataset's 'details' page — a key workflow improvement derived from our user research.

By supporting metadata sources, we've also made it much easier to manage metadata updates. As mentioned above, keeping metadata up-to-date can be complex. With these improvements, updating metadata becomes as easy as updating any ordinary data item on Koordinates (that is, pretty easy).  This is a significant workflow improvement — especially for organisations importing and managing large amounts of geospatial data.

Finally, we've also made it possible for data managers to ‘lock’ published data to the title, description, and tags contained in the XML metadata. This means that the official details of the dataset contained in the metadata will always be reflected in the published data. 

These improvements make life easier for data managers. But the ultimate beneficiary is the professional data user, who can export and use data with greater confidence in its providence and currency.