Page 271 - Special Topic Session (STS) - Volume 2

STS493 Irene S.

Figure 2: Example of the meta data model as a semantic representation of objects

Data lineage in turn makes it possible to trace which variables are used where
and in what outputs. When quality issues occur, it is possible to identify the
source of the error and which outputs are affected by it. A hot issue in a time
where apparently everybody seems to have access to massive amounts of data
are the fact vs fake discussions. Making it more and more necessary to file and
save processing steps in permanent and invariable data sources. Block chain
technology could be of great value for example in supporting third parties to
verify whether the data really originates from an NSI and the quality of the
processing steps. A process that starts at data collection and stops with the
dissemination.

2.7 Data scouting
Although registers and registrations may be all around, it is not self-evident
that CBS can have the data at its disposal. Firstly, it must be known that a data
source is available and secondly it must be determined whether this data source
is useful in making official statistics (whether or not combined with other data
sources). It is at this point of interchange between hard skills on data handling
and soft skills concerning relation management that CBS introduced the position
of Data scout. The aim of our Data scouts is to organize and facilitate the
acquisition and opening up of new (big) data sources in order to create new, or
improve/enrich existing statistics. Thereby a Data scout is the linking pin between
both internal and external stakeholders. Its portfolio comprises a wide range of
tasks varying between identifying the CBS data requirements together with
content experts, mapping possibilities of relevant new data sources, evaluating
and testing possible new data sources on their usability for the intended
purposes, working together with legal, technical and domain experts to discuss
the (im) possibilities, building new relationships with relevant partner
organizations, negotiating terms and conditions for data use, defining (joint)
business models and making agreements with data owners. Challenges that our
Data scouts encounter include the following areas: the need to anticipate
continuously on future possibilities (knowing possible applications before
knowing the data completely), long project lead-times, setting up joint business
models with private corporations, dependence on external parties, issues
concerning data sharing (Safety, access, storage, etc.) and various legal issues
(privacy, GDPR, collaboration with private partners etc.).

260 | I S I W S C 2 0 1 9