The European Union expands the categories to be considered high-value data

The concept of High-Value data (High-Value datasets) was introduced by the European Parliament and the Council of the European Union 4 years ago, in Directive (EU) 2019/1024. In it, they were defined as a series of datasets with a high potential to generate "benefits for society, the environment and the economy". Therefore, member states were to push for their openness for free, in machine-readable formats, via APIs, in the form of bulk download and comprehensively described by metadata. 

Initially, the directive proposed in its annex six thematic categories to be considered as high value: geospatial, earth observation and environmental, meteorological, statistical, business records and transport network data. These categories were subsequently detailed in an implementing regulation published in December 2022. In addition, to facilitate their openness, a document with guidelines on how to use DCAT-AP for publication was published in June 2023. 

New categories of data to be considered of high value  

These initial categories were always open to extension. In this sense, the European Commission has just published the report "Identification of data themes for the extensions of public sector High-Value Datasets" which includes seven new categories to be considered as high-value data  

•  Climate loss: This refers to data related to approaches and actions needed to avoid, minimize and address damages associated with climate change. Examples of datasets in this category are economic and non-economic losses from extreme weather events or slow-onset changes such as sea level rise or desertification. It also includes data related to early warning systems for natural disasters, the impact of mitigation measures, or research data on the attribution of extreme events to climate change. 

• Energy: This category includes comprehensive statistics on the production, transport, trade and final consumption of primary and secondary energy sources, both renewable and non-renewable. Examples of data sets to consider are price and consumption indicators or information on energy security.   

• Finance: This is information on the situation of private companies and public administrations, which can be used to assess business performance or economic sustainability, as well as to define spending and investment strategies. It includes datasets on company registers, financial statements, mergers and acquisitions, as well as annual financial reports.  

• Government and public administration: This theme includes data that public services and companies collect to inform and improve the governance and administration of a specific territorial unit, be it a state, a region or a municipality. It includes data relating to government (e.g. minutes of meetings), citizens (census or registration in public services) and government infrastructures. These data are then reused to inform policy development, deliver public services, optimize resources and budget allocation, and provide actionable and transparent information to citizens and businesses. 

• Health: This concept identifies data sets covering the physical and mental well-being of the population, referring to both objective and subjective aspects of people's health. It also includes key indicators on the functioning of health care systems and occupational safety. Examples include data relating to Covid-19, health equity or the list of services provided by health centers.  

• Justice and legal affairs: Identifies datasets to strengthen the responsiveness, accountability and interoperability of EU justice systems, covering areas such as the application of justice, the legal system or public security, i.e. that which ensures the protection of citizens. The data sets on justice and legal matters include documentation of national or international jurisprudence, decisions of courts and prosecutors general, as well as legal acts and their content. 

• Linguistic data: Refers to written or spoken expressions that are at the basis of artificial intelligence, natural language processing and the development of related services. The Commission provides a fairly broad definition of this category of data, all of which are grouped under the term "multimodal linguistic data". They may include repositories of text collections, corpora of spoken languages, audio resources, or video recordings.  

To make this selection, the authors of the report conducted desk research as well as consultations with public administrations, data experts and private companies through a series of workshops and surveys. In addition to this assessment, the study team mapped and analyzed the regulatory ecosystem around each category, as well as policy initiatives related to their harmonization and sharing, especially in relation to the creation of European Common Data Spaces. 

Potential for SMEs and digital platforms   

In addition to defining these categories, the study also provides a high-level estimate of the impact of the new categories on small and medium-sized companies, as well as on large digital platforms. One of the conclusions of the study is that the cost-benefit ratio of data openness is similar across all new topics, with those relating to the categories "Finance" and "Government and public administration" standing out in particular. 

Based on the publicly available datasets, an estimate was also made of the current degree of maturity of the data belonging to the new categories, according to their territorial coverage and their degree of openness (taking into account whether they were open in machine-readable formats, with adequate metadata, etc.). To maximize the overall cost-benefit ratio, the study suggests selecting a different approach for each thematic category: based on their level of maturity, it is recommended to indicate a higher or lower number of mandatory criteria for publication, thus ensuring to avoid overlaps between new topics and existing high-value data.  

You can read the full study at this link.