Abstract

While big data offers significant opportunities for research, policy, and practice, challenges remain in integrating such volume of detailed data for accurate and timely insights. To address these challenges, the Hungarian Central Statistical Office has initiated the Virtual Hungary pilot project, which seeks to organize the most detailed datasets into a fully integrated system. This enables the identification of relationships at an unprecedented level of detail. The overarching goal of the project is to formulate the digital statistical twin of the country by establishing a synthetic, nationwide socio-economic information system that provides continuous, quasi real-time insights for policymakers into the actual state of the economy and society. The Virtual Hungary project framework is planned to have the capability of linking data of different sources even at the individual level, while maintaining appropriate anonymity. Additionally, since data representation occurs at the elementary level, the information can subsequently be aggregated to any desired groupings or territorial levels. A major advantage is that it is now possible to examine and analyse cross-links between previously separate databases, while maintaining the highest level of data detail. This presentation aims to introduce the improved analytical potentials by selected examples such as examinations based on interlinked demographic, tax and corporate data. It will also be introduced how the project overcome on data imputation challenges by applying automated techniques, and also how the future aim of reducing the burden on data providers is reachable by the real implementation of the digital statistical twin.