This year, there has been made great progress in linked government data in Slovak Republic. And it was the most special year up to now. After many proposals, discussions and verifications, the concept of linked data was incorporated to the Strategic Priorities - Data and Open Data, that is part of the National Concept of Informatization of Public Administration (NKIVS). Additionaly, the first Central Ontological Model was approved (stable from 2016), but the best success is agreed policy for required interoperability level of public open data, i.e.: Public Money + Base Open Data = Linked Data. This formulation was refined many times to ensure it's feasibility 1) only new selected (strategical) data projects such as base registries need to publish linked data, 2) it is not necessary to use semantic database (triplestore); it is only required to use URI identifiers when data are published as open data or data are used in public services, 3) central open data publishing components for state and self-government administration will support RDF. For the sake of implementing linked data only to new projects, the checklist for new feasibility studies of information systems prescribes to declare support of interoperability through using URI & Central Model.
Since the linked government data being developed from 2013, the base semantic infrastructure has been fairly developed. The Central Meta-Information System (MetaIS) of Slovak Republic has been extended with subsystem for managing URI identifiers of slovak public resources in 2016.
To manage creation, usability and publishing of linked data, the methodology data.gov.sk-semanticweb is being developed in the data standardization group PS1. The methodology consists of around 50 topics that are divided to two main parts, the method and process. The method part defines for example - required URI format of public resource, the Central Ontological Model of Public Data (national ontologies derived from base registries), Central Model mappings to the open standards such as CPV, W3C Ontologies and others, rules for publishing open data (DCAT, ADMS, CPV), rules for required interoperability levels for public open data and many others. The follow examples show several selected topics:
- A.3. URI format of slovak public resource
The list of approved URIs in Central Meta-Information System
- A.3. Central ontological model of slovak public data
The Central ontological model contains national ontologies that were derived from base registries such as The Physical Person Register, The Legal Subject Register, The Address Register and others.
However, for the purpose of annotating published data with the ISA2 recommended standards, the Central Ontological Model is mapped to the Core Vocabularies, i.e. for example, national physical-person ontology is mapped to the Core Person Vocabulary
For the purpose of annotating published data, the DCAT-AP is used for describing catalogs, datasets a distributions and CPSV for public services. Therefore, since Central Model is mapped to the CPV and it is also used DCAT-ap, CPVS - it is possible to query slovak public data without slovak national ontologies.
- A.4.5.2 Rules for required interoperability level of public open data
The Strategic Working Group of NKIVS K9.4 Enhanced Data stated that: "every new informations system or component that will publish open data financed through public resources need to publish at least RDF (4* Open Data). However, if the that are highly reusable, for example open data from base registries, central codelists and similar, they need to be published as 5* Open Data. To be more precise, they need to be mapped to defined URI resources that are maintained in MetaIS (i.e. Central Ontological Model, registered URI identifiers for entities, etc.).
- A.4.5.3 Rules for publishing SK data catalog, dataset and distribution in RDF (DCAT, ADMS)
- A.4.5.4 Rules for publishing SK public services in RDF (CPSV)
For the purpose of publishing data about electronic services, the proposal of using CPSV for the National Agency for Network and Electronic Services was created
3.1 URI Registration
The URI Registration Process is managed by the Deputy Prime Minister’s Office for Investments and Informatization of the Slovak Republic (UPVII) with the Central Meta-Information system of Slovak Republic. Any government office can create a URI registration request for their desired resources, and the UPVII office do the registration with support of data standardization group. The MetaIS system provides follows type of URI registration: 1) URI for data element, 2) URI for dataset, and 3) URI scheme for group of resources. So, it is not necessary to register for example all URIs for all legal subjects, rather is is possible only to register URI namespace for all legal subjects as follows:
3.2 Central Ontological Model Administration
Current process of the Central Ontological Model administration is in competence of the data standardization group PS1 at UPVII. However, more general approach was introduced at the Semantics 2017 Conference: Slovak Public Metadata Governance and Management based on Linked Data:
that is transposition of the Methodology and tools for metadata governance and management for EU Institutions
4. LOD Slovakia
The LOD Slovakia is unofficial open project of linking public data that started in 2015. It is mainly used for analytical, presentation and validation purposes of slovak linked government data standards. The LOD contains key slovak base registries data and other key information systems represented as linked data. Each linked database holds just several example data, however, they can be used for full transformation. In this moment, LOD Slovakia is not linked to the global LOD, because the priority is to focus on active usability of linked data in slovak e-government. However, the LOD SK is planned to be incorporated to the global LOD (with mapping to the DBPedia), and when the time will come, we believe it will be a cherry-pie work.
Bratislava Old Town Muncipality
Ministry of Finance of Slovak Republic
Legal Subject Class
Slovak couties (LAU1) – Statistical codelist
4.1. Using Tripleskop to browse and query data
The Tripleskop is a web based visual client for a SPARQL endpoint which provides many functionalities to work with a RDF graph. The following image depicts selected properties of the Ministry of Finance of Slovak Republic.
Now it is possible to transform this graph to create valid SPARQL queries, that can be run against semantic database. When I remove unimportant properties and change desired nodes to variables and mark them to be in output, I can create a query that stands for: "Which organization in The Legal Subject Register with concrete legal form performs specific agenda and is situated in concrete street and district (in the Address Register)?"
Since the LOD Slovakia consists of many base registries and other linked public data, the Tripleskop is a great example tool for creating data mashup queries.
5. Upcoming challenges
The key success factor of using linked data in public systems is to secure staffing in public offices. However, the Deputy Prime Minister’s Office for Investments and Informatization of the Slovak Republic supported to establish "Data offices" on the central level and local levels by the end of March 2018. These data offices will be led by Chief data officers (SK Data Curator), that will manage several types of data experts for specified data related field, such as data quality/interoperability, data security and others. These roles and their activities will refine current data standardization process and working groups. We believe, that it is possible to successfully use linked data and get closer to the data driven state vision. More generally, in linked data we believe.
Dear friends and colleagues, let me say on the end that we will be grateful for any suggestion or comments.