Supporting Policy-Makers with Social Media Analysis Tools to Get Aware of Citizens' Opinions

1. Introduction

Governments and public institutions are increasingly working with citizens to give the citizens more of a stake in the policy-shaping process, for example through public consultations on new legislation (Koop et al., 2010). In practice e-participation platforms foster communication and interaction between politicians and government bodies on the one side, and citizens on the other. Notwithstanding the benefits brought about by existing eParticipation platforms, there remains the unsolved challenge of how to involve a larger number of affected individuals, groups and communities in discussions than is currently achieved through dedicated web sites. This problem has, for example, been analyzed with the 10 Downing Street Debate Mapper, being a case in point (Miller & Williamson, 2008). They found that very few people (7% of 309 invitees) took part using the dedicated Debate Mapper website, but a larger proportion did comment about the same subject on other Web platforms.

The use of social networking platforms has a significant part to play in political engagement. Social Media (Hrdinova et al., 2010) and blogs (Coleman, 2005) have high potential for the eGovernment to interact with citizens. From a sociological point of view, platforms like Twitter are interesting for analyzing the dissemination of topics and as well for analyzing the opinions and sentiments of the society regarding particular topics (Savage, 2011). Beyond that, online platforms have the power to influence the process of opinion making (Coleman, 2004). That’s why Coleman (2005) uses the label: “new politics of listening”. For instance in the UK the Internet and social networks are everyday life functionality for Parliamentarians (Williamson, 2009).

There is thus a huge potential of online discussion places, but there is a problem of making sense of the huge amounts of text in them. The aim of this paper is to suggest analysis tools and validation two cases to support eGovernment by exploiting the potential of online discussions and addressing the problem of “too much information”. In the next section, we introduce the EC IST FP7 WeGov project¹ which developed the WeGov Toolbox. The WeGov Toolbox (hereafter “the Toolbox”), will be explained afterwards. Then we introduce the toolbox analysis components that automatically determine key themes and opinions, key users and exemplary posts from comments from debates on social networks. Subsequently we explain the evaluation model behind these two case studies to validate the analysis results. Finally, we draw general conclusions.