Proposing a Three-Component Method for Mapping the Activated Public opinion: A Big Data Applied Analysis of the Greek Political Blogosphere

Abstract:

After 15 years of digital existence, blogs have always been a constant gateway to the multifaceted realm of Social Media Networks. Since their introduction, they have been among the most popular applications of Web 2.0 in the realm of politics, providing digital citizens with a user-friendly tool to post political content, mutually interact, shape the political agenda and horizontally influence the public opinion as well as vertically affect the administrative decision centers. Therefore, they have been envisaged as a reach Big Data field suitable for research over the so called Activated Public Opinion. According to Chadwick (2006), blogs refer to each other via hyperlinks, forming in that way their own social networks. In these networks, they tend to form clusters that correspond to communities within the overall political blogosphere. The aim of this paper was to introduce a combined qualitative as well as quantitative method for carrying out research over blogs with political content focusing on the formation of these blog communities, based on their hyperlink interconnectivity. The case study introduced in this paper focuses on the Greek Political Blogosphere. As a first step, 127 political blogs were recorded, which were found to be among the most popular in the survey period. Based on their hyperlinking properties, the adjacency matrix from the blogs’ social network was constructed. By applying Multidimensional Scaling and Cluster Analysis, clusters were located, which correspond to the aforementioned communities and consequently to the way political bloggers see the whole community they form. Taking a step further, research took place in order to investigate if political blogs, forming the above mentioned clusters, are somehow related to the degree they may potentially influence their readers. Findings showed that high influence does not always involve high rates of hyperlinking. Eventually, by applying  Content Analysis to the content of the blogs forming the clusters, the qualitative characteristics and the general topics of political discussion were identified, regarding to the  specific survey period. Findings suggested that the qualitative characteristics of the political blogs are significantly related to the cluster they belong to, according to their hyperlinking. In conclusion, a method of mapping the political blogosphere, consisting of three components, is presented in this paper. In the case study presented, the method proved to be able to investigate both its qualitative and quantitative characteristics. By applying it in the Greek political blogosphere, it became possible for the first time to investigate thoroughly the political debate that takes place on the internet, its characteristics and the influence it can have on public opinion.