Handbook of Computational Social Science for Policy

This open access handbook describes foundational issues, methodological approaches and examples how to analyse and model data using computational social science (CSS) for policy support. Up to now CSS studies have mostly developed on a small, proof-of-concept, scale that prevented from unleashing its potential to provide impactful and systematic impact to the policy cycle, as well as from improving the understanding of societal problems to the definition, assessment, evaluation, and monitoring of policies. The goal of this handbook is to fill this gap by analysing and modelling data for policy support. It is achieved by advocating the adoption of CSS solutions for policy and by raising awareness for existing implementations of CSS in policy-relevant fields. To this end, the book explores applications of computational methods and approaches like big data analysis, machine learning, statistical learning, sentiment analysis, text mining, systems modelling, and network analysis to different problems in the social sciences. The presentation is done in three steps, each one with its own part in the book: the first chapters on foundational issues open with an exposition and description of what the key policymaking tasks are in which CSS can provide insights and information. In detail, the chapters cover public policy, governance, data justice and other ethical issues. Part two consists of chapters on methodological aspects dealing with issues such as the modelling of complexity, natural language processing, validity and lack of data, or innovation in official statistics. Eventually, part three describes the application of computational methods, challenges and promises in various social science areas, including economics, sociology, demography, migration, climate change, epidemiology, geography, or disaster management. The book's target audience spans from the scientific community engaged in CSS research to policymakers interested in evidence-informed policy interventions, or private companies holding data that can be used to study social sciences and are interested in making a policy impact.