The Ontology of Tweets: Mixed-Method Approaches to the Study of Twitter

This chapter focuses on Twitter and the unique challenges associated with data collection and analysis on this microblogging platform. Specifically, many big data approaches that are popular for studying tweets are tremendously useful, but are often ill-suited to more in-depth contextualized analysis of tweets. The chapter speaks to this issue and proposes alternative approaches to create a more balanced means of analysis. Further, the chapter proposes a framework to categorize tweets, addressing issues of ontology and coding. It draws on qualitative approaches, such as grounded theory, to demonstrate the value of a solid coding scheme for the qualitative analysis of tweets. To illustrate the value of this approach, the chapter draws on a case study, the Twitter response to the controversial song Accidental Racist, and shows examples of how this emergent coding of Twitter corpora can be done in practice. This case study illustrates how the proposed approaches offer ways to tackle themes such as racism or sarcasm, which have been traditionally difficult to interpret. Finally, the chapter draws some conclusions around a) Twitter as a platform for mixed methods approaches, and b) the value of relying on established approaches, like grounded theory, to inform Twitter analysis.