CSS Group explores advanced methods of data science to monitor prevailing trends in a comprehensive set of well-being indicators, importantly combining both micro- and macro-level dynamics in an integrated multi-dimensional system. It utilizes broad, macro-historical theories and models based on the analysis of long-run patterns in historical data to run scenarios and forecasts of future dynamics. Recent research within the new discipline of Cliodynamics particularly the development of structural-demographic theory , has identified a number of processes that help us understand long-term trends in social dynamics. This macro-level information are combined with huge amounts of micro-social data concerning contemporary events ‘streaming’ into our analytic framework from both social (twitter, Facebook posts, etc.) and traditional media (online news outlets and aggregation sites). This will allow us to assess in real time when long-run dynamics leading to outbreaks of violence and instability erupt and to monitor how they spread.
CSS Group focuses on (1) collaborative theorization and analysis bridging the work of data scientists and social scientists from several fields. Further, it relies on (2) an innovative digital-analytic infrastructure now being developed for collecting and integrating socio-historical data. These two dimensions of the group are linked by several years of shared work between the Collaborative for Historical Information and Analysis (CHIA), and the Seshat Global History Databank. We aim at building a comprehensive information recourse on social well-being through a combination of historical and contemporary social information with simulation modeling based on social-science theory and on data science methods. CSS Group utilizes unique computational and data handling resources of Pittsburgh Supercomputing Center and scalable cloud infrastructure for the advanced information processing required for large-scale socio-historical data integration, data quality enhancement and data analysis.