Open data and linked open data

It is important to be clear about what it means to be open to ensure interoperability that when it comes to data, it is about the ability to interoperate or integrate different sets of data, allowing different components to work together. Thus, a part of the open material can be mixed with other open material, which is essential to understand the main practical benefits of opening: the dramatic increase in the ability to combine different databases or data sets and thus develop more and more. better products and services.

Access and exploitation of open data

The open data initiative, which has been joined by many organizations and public administrations, recommends exposing the data as soon as possible, without worrying much about the format. So it could be open data:

  • An Excel file or files (it is a proprietary format).
  • A file or files in CSV format (which have problems with texts with some format, such as jumps).
  • An API that provides access to disconnected tables or views.
  • Smart data. Data with information about the data.
  • Linked Open Data. Open data and linked (or linkable) with other data.

In other words, not all open data is the same, nor do they offer the same interoperation facilities. While it is true that speed is a factor of innovation, discomfort of use is undoubtedly a brake. In this sense, open data can have problems depending on the solution that is adopted:

  • Obtaining. Are files downloaded or is there an API to access the data?
  • Data model. Is it closed or open? Public or private? Standard or proprietary? Expressive? Extensible?
  • Relationship between the data. Is there an API on data views or is it possible to run queries with standard languages ​​(such as SQL or SPARQL)?
  • Data binding. Is the data relatable to external sources or is it closed?
  • Data update. Is the data downloaded or is there an API that always accesses the updated data? Is updating the responsibility of the reuser or the owner of the data?
  • Dissemination of the data. Are a series of links provided or is there a website that allows them to be consulted online? Is it a passive or active diffusion?

Data linked under the standards of the Semantic Web

Our solution to these problems goes through the use of Semantic Web standards, particularly those related to Linked Data. Tim Berners-Lee is responsible for both the Semantic Web concept - which he coined in 2001 in a seminal article published in 'Scientific American', 'Semantic Web' - and the Linked Data concept developed from a design note related to the construction of the Semantic Web project.

Linked Data or Linked or Linkable Data designate a method of structured publication of data that, in practice, is what makes it possible to navigate by hyperdata and the construction of Knowledge Graphs. This publishing method enables people to interrogate the data in a graph semantically.

To publish in accordance with the principles of the Linked Data Web or Linked Data Web it is necessary to use some standards such as http, RDF and URIs, which are used not so much to serve the pages that people read, but to edit the pages that They can automatically interpret the systems and therefore share the information with them. This makes it possible to connect the data from different sources in a unified and interrogated graph.

Therefore, Open Data projects with GNOSS are Linked Open Data projects, which can be quickly deployed and respond to possible problems with the highest level of interoperability:

  • Obtaining. They provide an API that allows access to the data of each of the entities individually.
  • Data model. The models are open, developed using standard ontologies for public use, extensible to add new entities or relationships, and with expressive data that inform people and machines of their meaning.
  • Relationship between the data. If the project requires it, they can provide a SPARQL interface for users to run their own queries on the data.
  • Data binding. Data can be linked to external sources and also be linked from external sources, contributing to the Linked Open Data Cloud.
  • Data update. Accessing the data through an API ensures that the reused data is always the latest, so that reusers do not have to worry about the data update process.
  • Dissemination of the data. The GNOSS open data projects have a web expression that allows the access and indexing of the data and its maximum dissemination through carefully designed websites and consultation tools that allow users