Information Dialogue: CBS answers your questions 24/7

/ Author: Miriam van der Sangen
© Sjoerd van der Hucht Fotografie
What if you could request data and other information from Statistics Netherlands (CBS) at any time of day or night? CBS is working on an application that should make this possible in the near future. The application is part of the CBS innovation programme ‘Information Dialogue’. The aim is to offer 24/7 information to policy-makers, companies and other users in the best and most accessible way.

Users want more

‘Here at CBS, we strive to disseminate and communicate about our research results using the latest methods,’ says Director of the Communication and News department Mike Ackermans. ‘We do so in multiple ways, including making statistical information available to the public in various open data formats. This is what we are working on around the clock. At the same time, we have seen demand change among the users of our information. More and more, users are expecting instant responses to their questions. In addition, the way in which users request information is changing with the arrival of digital assistants such as Google Assistant and Siri. We are tapping into this by introducing the Information Dialogue innovation programme. The application currently under development will soon be able to generate a reply to any question by any user at any time of day, with as much detail as possible.’

Prototype

Programme manager Gabriëlle van Mourik is closely involved with the development of the application, which is being designed to meet the user demand for direct information: ‘In May of this year, we started off with a small group of people in workshops. My colleagues Jolien Oomens and Sara Houweling tested various software packages for their suitability. After this, they built an application prototype which can be addressed by users to ask CBS questions in spoken or written form; in response, the application will produce an answer which is as specific as possible. A part of the provided information is retrieved directly from the existing open data and other figures available at CBS. At a later stage, the application will also be able to generate new open data from CBS microdata in response to a particular question. Obviously, we will make sure that no individual information on persons or companies is traceable when using such data.’

[video: https://www.cbs.nlen-gb/video/a4856a04493e4ad0a746de6a6bdeb6b7]

Ambitions

Up to now, application development has been in an experimental phase with the main focus on recognition of the questions asked by the user. A working proof of concept has already been created, based on a limited number of datasets. It is expected that CBS will launch an initial beta version in March 2019. The ambitions for the near future are high. Ackermans: ‘We want to work towards an application which can give an instant response to user questions based on all content available at CBS. This includes not only figures but also data visualisations, infographics, video messages etc. In this way, we offer our users an approachable way to access all our information.’

Innovation programmes

Together with the Advanced Data Collection and Big Data innovation programmes, the Information Dialogue innovation programme forms an integral component of the renewed innovation strategy which was launched by CBS at the end of 2016. Ackermans explains: ‘Advanced Data Collection is about developing innovative methods for data collection. The Data collection department are currently working on this project. Big Data is about analysing large amounts of data such as sensor data. This is taking place at the Centre for Big Data Statistics. The Information Dialogue is a project under the management of the Communication and News department. These three innovation programmes have been established in response to important changes taking place in society: the changing information needs of users, the implementation of Big Data in the production of statistics, and the combination of different data sources, resulting in swifter statistics production with a higher level of detail.’