Editors: Mirco Nanni, Costantino Thanos, Fosca Giannotti and Andreas Rauber
The information hidden within big data may provide the key to solving many problems within society, business and science. However, turning an ocean of messy data into knowledge and wisdom is an extremely challenging task.
In this paper, we outline our vision of big data analytics in Europe, based on the fair use of big data with the development of associated policies and standards, as well as on empowering citizens, whose digital traces are recorded in the data. The first step towards this objective is the creation of a European ecosystem for big data analytics-as-a-service, based on a Federated Trusted Open Analytical Platform for Knowledge Acceleration. The goal is to yield a data and knowledge infrastructure that offers individuals, scientists, institutions and businesses: (i) access to data and knowledge services, and (ii) access to analytical services and results within a framework of policies for access and sharing based on the values of privacy, trust, individual empowerment and public good. Several requirements need to be fulfilled at four distinct levels as discussed in detail below:
- Scientific and technological challenges. Solutions to difficult problems are needed, including: i) the development of new foundations for big data analytics, which integrate knowledge discovery from big data with statistical modelling and complex systems science, ii) semantics data integration and enrichment technologies, which make sense of big data and make them usable for high-level services, iii) scalable, distributed, streaming big data analytics technologies to master the intimidating volume and speed of big data.
- Data requirements. The potential value that lies within big data can only be unleashed if a proper, efficient, fair and ethical access to data is provided to the relevant actors. This poses many techno-social questions: who owns and who may use personal data? What is the real value of such data? How can different data sources be accessed and linked? How can individuals be empowered through the capability to access, use, handle and control the usage of, their own data? How can we communicate to individuals and institutions the social and/or economic impact of personal data? How can open data initiatives be boosted and federations of linked data developed?
- Education and data literacy. Acquiring wisdom and value from big data requires competent data analytics professionals with considerable expertise in statistics and machine learning. Skills must be developed on how to exploit data and their analysis to develop successful business initiatives. Moreover, given the pervasiveness of big data in most disciplines of human knowledge and research, elements of data science should be provided to students at all levels of education, from high-schools to university curricula.
- Promotional initiatives for data analytics and Big Data Analytics-as-a-service-as-a-service. In order for Big Data Analytics-as-a-service to flourish in Europe, we must have a strategy to promote development along some key directions. These include: supporting the creation of big data analytics centres that are accessible to researchers, public administrations, and medium and small companies; incentivizing the adoption of a layered framework to increase interoperability across the single data repositories; and European leadership in the development of privacy-enabling solutions.
Finally, a very effective way of promoting and helping the development of big data analytics is to create successful, large-scale showcases in high-impact application domains. Some particularly effective examples might include: smart cities and communities; big data analytics for developing countries; the management of the global market of jobs; the (quantitative) assessment of results of European projects and activities; and the development of Big Data-aware Official Statistics.
Final recommendations
Over the last 10 years, European research has invested significant capital in database and data mining technology, and has developed a strong base of expertise and innovation in these areas. Future actions should capitalize and advance on this base. To this end, our recommendations are:
- The EU should spawn federations of key public and private actors, in challenging multidisciplinary domains to provide a critical mass for starting up Federated Trusted Open Analytical Platforms for Knowledge Acceleration and creating incentives for further actors to join.
- The EU should support the creation of big data analytics centres accessible to researchers, public administrations, and medium and small companies.
- Funding and supporting the development of the technologies needed to empower citizens, public institutions and businesses based on the values of the Federated Trusted Open Analytical Platforms for Knowledge Acceleration.
- Promoting the development of a normative framework for the above mentioned empowering of citizens, public institutions and businesses along four dimensions: privacy-preservation, trust management, individual empowerment and public good.
- Promoting education of novel data scientists and ‘datacy’.
- Promoting incentives for providing data access to researchers, businesses and public administrations. Examples include: assigning rewards to and/or facilitating virtuous business actors that share and maintain open data portals; giving value to ‘data citation’ in research, i.e. recognizing the citations to a data collection as a valid bibliometrics indicator; and enforcing regulations.
Organization and methodology
A group of twenty researchers from the areas of core database technology and data analytics met in Pisa, Italy in May 2014 to discuss the state of research in these areas in the context of big data, its impact on practice or education, and important new directions. The attendees represented a broad cross-section of interests, affiliations, seniority, and geography.
Before the workshop, each participant submitted a short paper, to be shared among participants, summarizing his/her vision of big data analytics over the next 10 years. During the workshop two subgroups were formed: one focused mainly on data management issues introduced by big data and the other on the new analytical opportunities opened by big data. The two groups alternated between separate meetings/brainstorming sessions and plenary meetings to share results and keep the overall efforts focused. The workshop terminated with the collaborative preparation of a declaration of aims and an index of context that were later expanded into the present paper.
Link:
The White Paper is available for download at
http://www.ercim.eu/images/stories/pub/white-paper-BigDataAnalytics.pdf
Please contact:
Mirco Nanni, ISTI-CNR,
Tel: +39 050 6212843
E-mail:
ERCIM launched an initiative to identify emerging grand challenges and strategic research topics in Information and Communication Science and Technology (ICST).
In 2014, two initial Task Groups were formed to investigate topics related to “Big Data Analaytics” and “Cyber-Security and Privacy”respectively. Further topics are currently being identified. The groups were composed of researchers representing a broad cross-section of interests, affiliations, seniority, and geography. Their results are published in two White Papers, which are briefly presented in the following two articles.
The full version of both White Papers is available for download from
the ERCIM web site at http://www.ercim.eu/publications/strategic-reports
