The Linked Data Solution Landscape Project
Overview
There is currently no description of Linked (Open) Data solutions, which is comprehensive, user-friendly, extensible, accurate, and compliant with the Linked Data principles.
To overcome this issue the Publication Office of the European Union launched a project, with the support of the European Parliament's Preparatory Action - Linked Open Data in the European Public Administration. Its goal is to create a knowledge model and technical solution for Linked (Open) Data landscaping. The knowledge model, in the form of an Ontology, can be populated with collected solution description, in a way that promotes discoverability and interoperability. The author of the LDS Ontology is Linked Data Expert, Ivo Velitchkov, with contributions from Giorgia Lodi (STLab, CNR, Italy) and Pieter van Everdingen (PLND, Netherland)
Objectives
The criteria for success of the project were defined as:
- Comprehensiveness
For the solution to be comprehensive two dimensions are considered, number of aspects and number of described entities, in this case, Linked Data solutions. Regarding the aspects, this should be realised with a richer set of properties but also corresponding communication activities so that the community is active in maintaining their values.
Regarding entities, within the projects, an initial collection is created to describe a few solutions. The aim is to attract activity of the community and mainly the solution owners, which at one point in the future should bring the collection of described Linked Data solutions to to a number in the 4-figure range. - User-friendly
The user-friendliness includes both appealing user experience and inclusiveness – not requiring technical knowledge to search and interact with the catalogue.
The user experience covers both reading and editing (see User stories) - Extensible
Adding new elements should be made easy. Adding new properties should be embraced but subject to governance to ensure good quality of the model. - Compliant with the Linked Data principles
All 4 Linked Data principles should be complied with. The main care for the forth will be for the link between the solution description and the points of access to its services, as well as to meta-data describing the solution provider. - Accurate
The assurance of good data quality is a combination of built-in constraints (validation rules) and quality procedures.
Guiding Principles
The projects’ approach and solution architecture support the following descriptive principles:
- Interoperability
The proposed solution is coherent, interoperable, extensible, and allow independence from the future strategy or fate of a particular vendor. - Re-use
Re-use of components of useful and used ontologies, platforms, data visualisations, solutions, and technical approaches, is preferred over developing components with similar features. - Data-centricity
The solution design is data-centric. Both data and metadata are expressed in the same way. The semantics is explicit and based on a shared data model. The interpretation of the data is not in the application code or dependent on local database scheme. The shared data model (ontology) includes, where needed and possible, data validation and rules logic, developed in a modular way to allow Flexibility and support a variety of use cases.
Data-centricity, unlike application-centricity, supports the principle of user-centricity.
See also [Data-Centric Manifesto|http://www.datacentricmanifesto.org/] - EU data is EU asset
The provision of a solution, platform, coordination, gathering, curation, linking
or any other activities on a data asset does not lead to the possession or any
restriction of use outside personal data protection and compliance with defined security requirements. - Reciprocity
The engagement of stakeholders enables mutual benefits from the exercise.
Further Reading
- LDSO - Linked Data Solution Ontology
- Proof of Concept on EU Knowledge Graph - The ERA Route Compatibility Check Linked Data Solution described