Statistics, data science and national data management: three tasks for a forward-looking FSO

Statistical production, data management and data science Statistics, data science and national data management: three tasks for a forward-looking FSO

28.12.2020 - The Federal Statistical Office (FSO) has reorganised itself for the start of 2021 to meet the challenges that tomorrow will bring. To go beyond its key mission as a producer of statistical information and to comply with various decisions made by the Federal Council, the FSO has created a division dedicated to the national management of data and a new national data science competence centre.

Implementation of the National Data Science Competence Centre (DSCC) is underway so that the FSO can work on a new task entrusted to it by the Federal Council. The competence centre will be incorporated into the new Data Science and Statistical Methods division, headed by Bertrand Loi-son.

The DSCC will operate as a service provider, offering "Data Science as a Service" (DSaaS) services aimed at solving complex, unstructured and data-rich problems by using technical and practical methods specific to data science, arising from areas such as machine learning and artificial intelli-gence. In the long term, the centre aims to acquire as much knowledge as possible through con-crete projects. These new services will be available to the entire federal administration, in coopera-tion with the cantonal and communal authorities.

"The FSO is evolving and becoming a place for experimentation while respecting data protection at all times. The idea is that anyone within the federal administration with a data science question can call and ask us if anyone has had a similar problem and whether a solution was found or a new one is needed," explains Georges-Simon Ulrich, Director General, indicating that this is an example of the services that the DSCC will offer in future. See his detailed explanations on video.

Interconnected within the federal administration, the competence centre will benefit from the experience of a scientific team who all have an excellent academic background in the fields of mathematics, statistics, data science and data engineering. "The ultimate goal is to make data more visible and usable, to improve our analyses and to gain a better understanding of what is going on around us".

Reusable data

The FSO’s second new task concerns the data themselves. Headed by Manuela Lenk, the new Interoperability and Registers (IOR) division is developing Switzerland’s data ecosystem. Data from the federal administration needs to be easier to use in order to avoid asking businesses or private households for the same information several times. To achieve this goal, the FSO also intends to manage a precise inventory of the contents and data collected by the entire federal administration.

This information, known as metadata, will be compiled in a centralised catalogue, making them easier to consult and use with the help of algorithms from the field of data science. This knowledge base will make it easier to reuse data many times.

Furthermore, in order to make these data reusable, norms must be established so that all the systems can communicate between one another: this is known as interoperability. Data must be produced within a carefully documented framework and be machine-readable in order to be exchanged easily.

The FSO’s key mission remains

Alongside these two new directions, the FSO’s basic mission, which consists of producing reliable, high-quality statistical information, remains. In our digital information society, this mission is more important than ever.

As has always been the case, the FSO’s statistical results must enable the public to form an opin-ion and contribute to political and economic decision-making. Produced independently and trans-parently according to international good practices, this statistical information allows us to gauge the Switzerland of today and to compare it with the rest of the world.