Comparing automated content analysis methods to distinguish issue communication by political parties on Twitter

Party competition in Western Europe is increasingly focused on “issue competition”, which the selective emphasis issues by parties. The aim of this paper to contribute methodologically increasing number studies that deal with different aspects parties’ issue and communication. We systematically compare value shortcomings three exploratory text representation approaches study communication parties Twitter. More specifically, we analyze separate online one party from other how consistent is. Our analysis was performed two years Twitter data six Belgian political parties, comprising over 56,000 tweets. results indicate our approach useful profile themselves strategies are at play. Second, method allows individual politicians contributes classical literature unity discipline. A comparison methods shows a clear trade-off between interpretability discriminative power, where combination all simultaneously provides best insights.