Public Sector Experiments with Social Media Data Mining

Abstract

Kennedy discusses action research with city-based, public sector organisations, which attempted to experiment with these methods, evaluate their potential use and reflect on their normative consequences. She argues that it is difficult to discuss the problems of data mining openly with public sector actors because doing social media data mining is motivated by a will to produce results, or a ‘desire for numbers’—this concept brings together Porter’s discussion of trust in numbers (Trust in numbers: The pursuit of objectivity in science and public life. Princeton, NJ: Princeton University Press, 1996) with Grosser’s (Computational Culture: A Journal of Software Studies, 4, 2014) work on the ways in which the metrification of sociality on social media platforms creates a desire for more and more metrics. Nonetheless, public sector social media data mining aspires to enhance understanding of public opinion and inclusion in public processes and so aims to serve the public good. Kennedy argues that it is therefore ‘empirically inaccurate’ (Banks, The politics of cultural work. Basingstoke, England: Palgrave Macmillan, 2007) to understand public sector data mining in only negative terms.