Page 146 - Invited Paper Session (IPS) - Volume 1

IPS122 Elise C. et al.
If the previous conditions are fulfilled, then a mixed SSP Lab/Business
Unit/IT team with the required skills (datascience, IT, etc.) is set up. Flexibility
and real commitment to the project are required from the different
participants. Several models of engagement, e.g. one day-per-week
contractual commitments, customised projects and team descriptions are
possible. Flexible and ‘agile’ ways of working are promoted. The mixed team
is expected to work in steps and cycles, and to deliver regular outputs. The
requirement is to broadly disseminate the results (final and intermediary) and
to share experience through different media (intranet, newsletters, etc.).

b. Networking activities
For some experimental projects, the SSP Lab contributes to and benefits
from European expertise. In particular, the SSP Lab takes part of the European
exploration of the potential of Big Data to integrate it into the official statistics
production (Big Data task force and related ESSnets). It participated in the
ESSnet Big Data I Mobile Data Working Group, which aimed to 5 clarify the
possibilities of accessing mobile phone data, lay the foundations for a
methodology for their treatment and estimate population present within a
given place and time and related indicators. The SSP Lab will continue to
participate in the investments on mobile phone data in ESSnet BD II and will
coordinate the French participation in other work packages, including those
on satellite data and smart statistics.
For other experimental projects, the SSP Lab may also form academic
partnerships in order to benefit from external expertise. For instance, it
currently collaborates with the Institute of Public Policies (IPP) of the Paris
School of Economics (PSE) to explore the modelling of professional careers for
microsimulation purposes by using machine learning methods. Partnerships
with private actors for experimentation on private data are also in the scope
of potential activities.

c. Dissemination
The SSP Lab ensures the role of monitoring and dissemination of
innovative statistical methods through training on datascience methods (e.g.,
machine learning, textmining and coding languages such as Python) for the
statisticians of the official statistics service and the provision of technical
documents. The SSP Lab animates networks on innovative topics within the
SSP (dissemination of a Big Data newsletter and Big Data seminars). The SSP
Lab and the IT innovation unit also work together to promote 'agile’ ways of
working, in particular via collaborative workshops and hackathons open to the
members of official statistics service and close institutions.

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