This article was originally published on Cities-Today.com
By Bart Rosseau, Chief Data Officer, City of Ghent
Linked open data
Linked open data on a semantic website might seem distant from the daily operations of local government, but in Ghent we believe the benefits outweigh the possible pitfalls of using this pioneering technology.
Originally defined by Sir Tim Berners Lee, the semantic web aims to connect data and meaning (not just the values but also the definitions) on a scale the size of the Internet. So you have not only a web of text and images but also of data with the same ease to jump from one data point to another, enabling artificial intelligence and humans to reason over these data points. Linked open data on a semantic website might seem distant from the daily operations of local government, but in Ghent we believe the benefits outweigh the possible pitfalls of using this pioneering technology.
Using the web as a distribution platform of contextualised data will allow us to distribute content, and filter datasets that are disconnected by design. Through machine-readable definitions and uniform resource identifiers (URIs), government-managed datasets become anchor points, allowing external data providers to add their content to those concepts. This allows for a more extensive knowledge graph based on linking data to content, rather than coupling on a technical level. As this is distributed over the Internet, there is no limit to possible links with other cities, government levels, companies, civil society and others.
Most of all, in Ghent we have experienced the benefits in a more internal alignment. By developing the semantic vocabularies we saw different city services agreeing on common terminology and meaning, and agreeing to coordinate data governance.
More institutions are adapting this technique, and the need to recognise and attract people with the right skillset is growing. Semantic experts and linked-data professionals are not knocking on our door (yet) but as more open source tools and accessible training programmes become available, the understanding, added value and sustainability will become essential to any data governance unit.
There is a huge difference in this evolution compared to other trends. The DNA of the semantic web is to reference data to validated concepts (ranging from peer-to-peer validation to internationally approved standards). There is a need to (re)use existing definitions, but also provide enough space to expand the definitions to suit specific needs.
Until now, two projects have proved to be milestones in making us acquainted with the technology and its possibilities.
The first project was internal, where we added a machine-readable semantic context to our existing webpages. This meant that a published news item on our website became enriched with information on the location it referred to–the relevant city department, the responsible politician, its theme and so on. By applying the right query language, a local startup could filter the information and distribute it on their hyper local platform.
The OASIS project (https://oasis.team/) is a good example of how we can work internationally. The project (supported by the EU through its Connecting Europe Facility programme) has enabled us to work with Madrid to define vocabularies for public transport and city-issued government services.
By adopting defined vocabularies, a web-based query to find government services, opening hours, the right department and the relevant public transport to take you there, becomes available, regardless of the language the web-based text is published in.
Of course, there are still some challenges. The short-term budgetary gains are not apparent, so the business case requires some more long-term thinking. There are still some debates on what supporting technology will prove to be the industry standard. The tools to make the wealth of information available to non-technical people are eagerly awaited.
The deep understanding and practical value needs to be developed, to embed this into the existing practices of IT and organisational development.
However, the implications and possibilities exceed the traditional boundaries of data management. Defining concepts and providing validated URIs for key policy tools and identities means that the role of the government as a validating, trusting authority becomes articulated in a digital world. For instance, providing URIs to government-approved initiatives–schools, philanthropic organisations and others–will add credentials that are easily shared and used in a digital environment.
Defining concepts by different government levels provides a more solid core to organise intergovernmental reporting and data exchange.
We are still in the early stages of this technology, the challenges are there, but so are the benefits.