Determining Citizens’ Opinions About Stories in the News Media

Abstract

We describe a method whereby a governmental policy maker can discover citizens’ reaction to news stories. This is particularly relevant in the political world, where governments’ policy statements are reported by the news media and discussed by citizens. The work here addresses two main questions: whereabouts are citizens discussing a news story, and what are they saying? Our strategy to answer the first question is to find news articles pertaining to the policy statements, then perform internet searches for references to the news articles’ headlines and URLs. We have created a software tool that schedules repeating Google searches for the news articles and collects the results in a database, enabling the user to aggregate and analyse them to produce ranked tables of sites that reference the news articles. Using data mining techniques we can analyse data so that resultant ranking reflects an overall aggregate score, taking into account multiple datasets, and this shows the most relevant places on the internet where the story is discussed. To answer the second question, we introduce the WeGov toolbox as a tool for analysing citizens’ comments and behaviour pertaining to news stories.  We first use the tool for identifying social network discussions, using different strategies for Facebook and Twitter. We apply different analysis components to analyse the data to distil the essence of the social network users’ comments, to determine influential users and identify important comments.