API, chatbot & taxonomy study!

The reports on automating data exchange through APIs, the architecture of public service chatbots and the European taxonomy of public services have been published!


All reports include a section on how these technologies can support the Single Digital Gateway and the role of a standard data model like CPSV-AP in using efficiently those technologies in the public administration’s context.

The ISA² action "Catalogue of Services" (CoS) is part of a European Programme to support member states to create interoperable solutions for public services and to support the Single Digital Gateway (SDG). The Single Digital Gateway is a European regulation requiring all member states to share information at the EU level concerning a set of public services. In light of this, the CoS action has worked intensively over the last months on updating the CPSV-AP, its Core Vocabulary for describing public services, and for identifying potential use cases to support some policy areas, including the SDG.

In parallel, we have explored how new technologies could be useful, together with well-structured vocabularies like CPSV-AP to create national and cross-border catalogues of services.

Report 1 looks at how public administrations should go about automating data exchange using APIs and what important considerations need to be factored in. You can find the report here.

APIs are increasingly needed as the demand for transparency regarding the data provided and the data consumption by the public sector increases. This challenges traditional data management processes. Nowadays, citizens and businesses expect public administrations to provide a fast and secure access to data while establishing proper connections with other (cross-border) public administrations and public services.

As building an API is no small feat and represents a journey for which public administrations are often underprepared, the report analysed the usage of APIs in the public sector, discusses safeguards and looks at key aspects that administrations should consider.

The report also explored how a common data model like the CPSV-AP makes it possible to gather descriptions of public services with APIs in an interoperable way, i.e. the descriptions collected from API could be crossed and reused together with the descriptions from another API.

Report 2 explores how citizens and business can best interact with public administrations to access  info on public services via chatbots. Those chatbots would benefit from making use of machine readable data models. You can find the report here.

With the exponential growth of AI technologies, the wish of public administrations to work more efficiently and the growing demand of citizens requesting a smoother delivery of public services, there has been a rise in the provision of public service chatbots. This report includes an analysis of existing public service chatbots and their costs and benefits. The report also provides a high-level architecture and a more detailed analysis of public service chatbots, by answering questions about the look, the personality and the medium of the chatbot, and many more.  

From our analysis and experience, we observe that the maturity of a chatbot increases with the volume and quality of the data used to train it. In this realm, this report also looks at how a data model like the CPSV-AP can be used to structure and improve your data quality. Moreover, the CPSV-AP can be used as a solid basis to build chatbots on: it links different information concerning public services that is important for the development of a chatbot.

Building a chatbot is not an easy walk but this report can support public administrations as it combines the major building blocks of a public service chatbot with an overall view on chatbot security and governance.

Report 3 identifies how a commonly agreed taxonomy of generic public services could be implemented to help public administrations harmonizing their catalogues of services. You can find the report here.

The last report provides the user with a detailed context of taxonomies. It provides insights into why using taxonomies is relevant and what kind of taxonomy could be used in a specific scenario. This report analysed that taxonomies should be used in conjunction with machine-readable data models like the CPSV-AP: the public services delivered by different organisations will be described in a harmonized way.

The report includes a proposed first version of a taxonomy for public services, which could be used to categorise public services. In the report, it is explained what existing information and methodology was used to create this taxonomy.


If you would like to get in contact with the team behind the reports, please contact miguel.alvarez-rodriguez@ec.europa.eu.

Shared in

Login or create an account to comment.