Page 119 - ITU Journal - ICT Discoveries - Volume 1, No. 2, December 2018 - Second special issue on Data for Good

ITU JOURNAL: ICT Discoveries, Vol. 1(2), December 2018

a) Data sovereignty d) Data exchangeability

Data sovereignty allows companies, data owners, to Data exchangeability ensures that data can be
keep control over their data. It is important for exchanged between a data producer and data
business to enter the data market with their consumer in general and be used for target
proprietary business data and be confident that applications or intended purposes. To much extent,
their data is not compromised or used by third this implies data format and platform and APIs
parties, without consent from the data owner (or compatibility, which is achieved by industry
data controller). Sovereignty is a key principle of the standardization. Ideally, data exchange should be
Industrial Data Space Architecture as defined by possible between compatible processes or
International Data Space Association [18]. applications during the whole data processing flow
or data lifecycle. In the context of economic data
Often the data sovereignty principle is opposed to value, exchangeability of data can also mean the
general data storage and processing on clouds possibility of exchanging data for other economic
where data resides in the cloud provider’s data goods or money. However, data pricing models are
centers and there is fear that companies may lose not addressed by the authors at this stage, although
control over their data, or the cloud provider may some references to related works are provided
have unauthorized access to data or their use for below.
business purposes. However, modern cloud and e) Data actionability
infrastructure virtualization technologies, provider
business models and compliance provide a When data is purchased by companies they should
sufficient number of controls to satisfy security and serve their business purposes and contain the
trust requirements by companies to operate their necessary information to derive actionable
businesses and host data on clouds (see CSA decisions about operation or process optimization,
Complete Cloud Security Governance, Risk, and in particular customer experience improvement or
Compliance (GRC) Stack, Cloud Security Alliance quality of services delivered. When data is used in
[23]). industrial processes, the actionable data must be

b) Trusted data extracted and included in the industrial processes
control. With increased use of artificial intelligence
Using data in decision making or in the processes in industry, the spectrum and variety of data used in
control requires that data is trusted and verifiable. the industrial production value chain is increasing
Trust in data is achieved by the whole process of and may include process monitoring data, logistics
data collection and by using verified models of the data, market data and user feedback data.
processes that data represent, which must be in f) Data measurability
general auditable. In most business cases, data
trustworthiness is ensured by the reputation of the Data measurability can be discussed at least in two
data provider but all aspects of data production and aspects: as an important property for data valuation
origin must be verifiable and auditable. and exchange as economic goods, and as a part of
c) Data reusability data handling on the data infrastructure platforms.
The former still requires additional research and
Data reusability should allow multiple uses of data, effective data valuation models, yet to be created.
even if it is not for the original purposes that the The latter is concerned with the resources that are
data was created for. Normally, data represents required or consumed by the data storage or data
events, entity or processes and are application processing facility or applications. This area is well
agnostic. Data reusability can create multiple developed and supported by modern cloud based
opportunities for data economy actors, including data infrastructure. Both public cloud platforms
small and medium enterprises (SME) or individual (such as AWS or Microsoft Azure) and Open Source
researchers. Data reusability should be supported platform (such as OpenStack or CloudStack)
by well-defined metadata and well-documented provide a rich opportunity to monitor cloud
data collection processes. Data reusability is part of resources usage by all data handling processes with
the research data FAIR principles and is well details up to processor cycles, storage transactions
supported by metadata management tools [19].

© International Telecommunication Union, 2018 97