A framework for evidence based policy making combining big data, dynamic modelling and machine intelligence

Governments and policy makers are striving to respond to contemporary socio-economic challenges, however, often neglecting the human factor and the multidimensionality of policy implications. In this chapter, a framework for evidence based policy making is proposed, which integrates the usage of open big data coming from a multiplicity of sources with policy simulations. It encompasses the application of dynamic modelling methodologies and data mining techniques to extract knowledge from two types of data. On the one hand, objective data such as governmental and statistical data, are used to capture the interlinked policy domains and their underlying casual mechanisms. On the other hand, behavioural patterns and citizens' opinions are extracted from Web 2.0 sources, social media posts, polls and statistical surveys. To combine this multimodal information, our approach suggests a modelling methodology that bases on big data acquisition and processing for the identification of significant factors and counterintuitive interrelations between them, which can be applied in any policy domain. Then, to allow the practical application of the framework an ICT architecture is designed, with the aim to overcome challenges related with big data management and processing. Finally, validation of the approach for driving policy design and implementation in the future in diverse policy domains, is suggested.