According to a European Data Market study[1], enterprises are increasingly aware of the relevance of digital business models as a critical success factor. They also clearly perceive the relevant role of data assets and the capability to manage internal and external data efficiently and effectively for strategic objectives.
Establishing sustainable data-based ecosystems has been one of the main objectives of the Big Data Value Public Private Partnership (PPP) (between the EC and BDVA, under the programme H2020), and the ultimate goal of projects funded under this portfolio was to contribute to such objective. Starting back in 2017 with, among others, large-scale pilot projects in domains of strategic importance for EU industry[2] and key ecosystem enablers[3], this has since continued with advanced platforms, tools, and testbeds (cloud, HPC, IoT) in emerging domains[4]. With the last projects under this programme focused on data platforms[5] and experimentation, incubation and a network of data innovation hubs[6], the EC has paved the way for a robust data European ecosystem.
The European Data Strategy published in 2020 leveraged assets produced by PPP projects to create a European Common Data Space, which seeks to connect stakeholders on a cross-border and cross-sector basis and share data in a trusted, secure and scalable way. A new tranche of projects funded through the Horizon Europe programme and under the umbrella of the ADR PPP is focused on producing the tools and building blocks needed to build data spaces, among them technologies and solutions for data trading, monetisation, exchange and interoperability[7], advanced technologies for data management[8] and, more recently, the integration of data life cycle, architectures and standards for complex data cycles and human factors[9]. Other projects address interoperability challenges in specific domains (e.g., energy[10]). Additionally, projects under “Extreme data mining, aggregation and analytics technologies and solutions”[11] are producing solutions to cope with “extreme data” (increasing volume, speed, variety; complexity/diversity/multilingualism of data; the dispersed data sources; sparse/missing/insufficient data/extreme variations in values), that can be applied when current technologies fail.
At the same time, projects funded under the Digital Europe programme (including preparatory and deployment actions) focus on establishing data spaces in relevant sectors (the DSSC project includes them under its community of practice[12]). The DSSC is working on producing a blueprint[13] (including a glossary, conceptual model, building blocks and collection of standards) to be adopted by data spaces designers and builders. This blueprint will include the federation of data spaces based on identified synergies. Finally, the SIMPL project aims to provide an open-source, intelligent, secure middleware platform supporting data access and interoperability among European data spaces[14].
Regarding the sustainability of data spaces, the European Digital Infrastructure Consortium (EDIC), an instrument proposed by the EC under the Digital Decade programme to speed up and simplify the setup and implementation of multi-country projects, has been the path followed by some member states to take the results from some EU data space projects and continue this work towards a broader scope of applicability and sustainability[15].
To complement the above, the EC recently (January 2024) published a second version of the Staff Working Document on Common European Data Spaces[16], which provides an overview of the current status.
In AI ecosystems, the DeployAI project will continue supporting the European AI-on-demand platform[17] to facilitate access for SMEs and the public sector.
This investment has come together with relevant pieces of regulation (by the European Data Strategy[18]), such as Data Act, Data Governance Act, free flow of non-personal data regulation, AI Act and GDPR. The regulation related to the European Health Data Space[19] is also worth mentioning.
As considered in the Data Governance Act, the EC has established the European Data Innovation Board (EDIB)[20], with representatives from member states, the EC, the European Union Agency for Cybersecurity, SMEs, and other bodies, to facilitate the sharing of best practices on data intermediation, data altruism, the use of public data, and the prioritisation of cross-sectoral interoperability standards.”
Finally, the Open Data Directive[21] promotes fair access to information and increasing the number of available public sector datasets, and at the same time motivates the public and private sector to reuse data. A High Value Dataset is a dataset holding the potential to (i) generate significant socio-economic or environmental benefits and innovate services; (ii) benefit a high number of users and SMEs; (iii) assist in generate revenues; and (iv) be combined with other datasets. According to this, the EC is supporting under the Digital Europe Programme a portfolio of projects to promote the use of high-value datasets[22].
Standardisation efforts are gaining high relevance in this field, given the importance of defining and agreeing on trust, interoperability and governance levels. The CEN CENELEC Workshop on Trusted Data Transactions[23] aims to anticipate standardisation requirements by defining terminology, concepts, and mechanisms for common elements that can compose a potential framework under which trusted data transactions can be based independently of any architectural choices or technical implementation. The recently created CEN CENELEC Focus Group on Data, dataspaces, cloud and edge[24] focuses on developing a standardisation landscape, identifying gaps and providing recommendations to Technical Committees on priority areas for standardisation in data, dataspaces, cloud and edge. The CEN CENELEC JTC21[25] has received the request from the EC to provide a standardisation plan for implementing the AI Act. Although firmly focused on AI, JTC21 also covers aspects like data quality, data governance and trust framework relevant to this theme. The W3C consortium also includes different Work Groups (WGs) addressing pertinent elements that can support various levels of interoperability, e.g.[26], the Dataset Exchange WG, the Decentralized Identifier WG, the Verifiable Credentials WG, and the ODRL community group. Regarding interoperability, it is worth mentioning the initiative promoted by the EC to foster the semantic interoperability of interconnected e-government systems (Semantic Interoperability Community Europe SEMIC), and the work of some key European projects like INTERCONNECT[27] and its Semantic Interoperability Framework (SIF).
Considering all the above, BDVA already identified in the paper “Towards a European Data Sharing Space: enabling data exchange and unlocking AI potential[28]” published in 2019, the need for data spaces, platforms and marketplaces as enablers to unleash the full potential of data at EU level to develop a powerful European AI technology according to European Values. This document was complemented in 2020 with a second version highlighting the opportunities behind all the previously identified challenges. Finally, in the paper “Data sharing spaces and interoperability”[29] published in 2023, BDVA describes a value perspective of interoperability using the value wheel presented in the previous papers, also providing practical guidelines to achieve meta-data interoperability and a potential roadmap.
On the private side, it is worth mentioning how BDVA, Gaia-X, FIWARE, and IDSA have joined efforts in the Data Spaces Business Alliance to foster the adoption and use of data spaces worldwide. In addition to the approach followed individually by each association, they are working together to achieve technology convergence, as witnessed by the DSBA Technology Convergence discussion document[30].
All the previous efforts are conditioned or reinforced by some general trends that are driving the whole data landscape:
- Implementation of ecosystem interoperability at different levels. It is worth mentioning, among others, the European Interoperability Reference Architecture[31] and other different efforts like the mentioned SIF[32].
- Different standardisation efforts related to data spaces and operations therein are taking place as a key aspect to enable scale-up and wider adoption. In addition to the ones previously presented, it is worth mentioning the new ISO/IEC AWI 20151 (Dataspace concepts and characteristics) under ISO/IEC JTC 1/SC 38[33].
- AI is used at different levels to support managing and operationalising many activities in data spaces, particularly[34], and digital ecosystems in general. Among those aspects, it is relevant to note how LLMs are being investigated for their application in providing data semantic interoperability.
- Management of digital identity in large ecosystems. For example, the last European Digital Identity regulation (eIDAS and eWallet[35]) was agreed upon last Nov 2023. In this regard, decentralisation in the governance and management of large ecosystems is the way to enable scalability and flexibility (although it also poses other risks and disadvantages). Decentralised identifiers (and their use in verifiable credentials) and self-sovereignty identity (SSI) are proposed by DSBA and Gaia-X, respectively, to manage digital identities.
- As people become more aware of the value that their personal data has and the different possibilities that appear (also relying on GDPR to protect their rights), there is increasing interest in individuals sharing their personal data and directly participating in data spaces and expecting some value out of it.
- The Data Governance Act introduces the role of “data intermediation services providers (DISP)”, which is gaining high relevance in data ecosystems because it could affect many of their activities, being the regime for DISPs not a voluntary scheme, rather a mandatory set of obligations, whose compliance may require a business to restructure its operations fundamentally.
- Support provenance and traceability, so that products and data can be tracked through the value chain and across different ecosystems. For example, the EU Digital Product Passport[36] aims to increase transparency, sustainability, and recyclability.
- Definition, negotiation, management and enforcement of contracts and policies to rule and enforce access to data, consent management, access rights, etc …
- Appearance of specific techniques to address data scarcity (small, rare, costly/unbalanced) or lack of data (e.g. data augmentation, transfer learning, synthetic data generation).
The deployment of more applications on top of data spaces and the resulting value delivery. This implies the development of relevant use cases and novel data-driven business models and designing and implementing the corresponding technical enablers to materialise them. This trend includes, as a paradigmatic example, the leverage of data spaces and all their full potential to power AI and fuel the training of (large) models
[1] https://digital-strategy.ec.europa.eu/en/library/results-new-european-data-market-study-2021-2023
[2] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/ict-15-2016-2017
[3] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/ict-17-2016-2017
[4] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/ict-11-2018-2019
[5] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/ict-13-2018-2019
[6] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/dt-ict-05-2020
[7] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/horizon-cl4-2022-data-01-04
[8] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/horizon-cl4-2021-data-01-03
[9] https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/horizon-cl4-2023-data-01-02
[10] https://interconnectproject.eu/
[11] https://cordis.europa.eu/programme/id/HORIZON_HORIZON-CL4-2022-DATA-01-05/en
[12] https://dssc.eu/space/DC/27983886/Community+of+Practice
[13] https://dssc.eu/page/knowledge-base
[14] https://digital-strategy.ec.europa.eu/en/policies/edic
[15] https://language-data-space.ec.europa.eu/related-initiatives/alt-edic_en
[16] https://digital-strategy.ec.europa.eu/en/library/second-staff-working-document-data-spaces
[17] https://www.ai4europe.eu/
[18] https://digital-strategy.ec.europa.eu/en/policies/strategy-data
[19] https://health.ec.europa.eu/publications/proposal-regulation-european-health-data-space_en
[20] https://ec.europa.eu/transparency/expert-groups-register/screen/expert-groups/consult?lang=en&groupID=3903
[21] https://eur-lex.europa.eu/eli/dir/2019/1024/oj
[22] https://digital-strategy.ec.europa.eu/en/news/eu-funded-projects-using-high-value-datasets
[23] https://www.cencenelec.eu/news-and-events/news/2023/workshop/2023-02-10-data-transactions/
[24] https://www.cencenelec.eu/news-and-events/news/2024/brief-news/2024-02-16-dataspaces/
[25] https://www.cencenelec.eu/areas-of-work/cen-cenelec-topics/artificial-intelligence/
[26] https://www.w3.org/groups/wg/
[27] https://interconnectproject.eu/
[28] https://bdva.eu/download/92/publications/3142/towards-a-european-data-sharing-space-position-paper.pdf
[29] https://bdva.eu/download/92/publications/3841/data-sharing-spaces-and-interoperability-bdva-discussion-paper-december-2023.pdf
[30] https://data-spaces-business-alliance.eu/dsba-releases-technical-convergence-discussion-document/
[31] https://joinup.ec.europa.eu/collection/european-interoperability-reference-architecture-eira
[32] https://interconnectproject.eu/resources/semantic-interoperability-framework-sif/
[33] https://www.iso.org/standard/86589.html
[34] E. Curry, et al., “Foundation Data Space Models: Bridging the Artificial Intelligence and Data Ecosystems (Vision Paper),” in 2023 IEEE International Conference on Big Data (BigData), Sorrento, Italy, 2023 pp. 190-195.
[35] https://ec.europa.eu/commission/presscorner/detail/en/ip_23_5651
[36] https://www.europarl.europa.eu/committees/es/digital-product-passports-enhancing-tran/product-details/20220510CHE10181