Characterizing issue fields on Twitter
emlyon faculty: Clément Levallois and Jean Savinien
collaborators: Mohamed Benabdelkrim (PhD student, emlyon & Univ Lyon 1), Cécile Robardet (Professor, Univ Lyon 1).
time frame: 2017-2020.
In this project, we make a methodological contribution to the empirical investigation of fields, by providing a principled procedure to map the boundaries, actors and structure of issue fields, at scale. The method leverages the classification acts performed by users of social networks who create lists of fellow users, which amounts to a social engine of classification or curation. Taking Twitter as a prime use case, our methodology aggregates connections between Twitter users based on their co-memberships in lists to create a map of an issue field, starting from a handful of seed Twitter accounts. Using network analysis (primarily, community detection) and semantic analysis, we characterize the identity and structure of the field. We test the method by replicating a previous study of the field of “social impact of nonprofits” and show how the map produced by the method can contribute to operationalize mechanisms at play in the field. We conclude by noting that the principles of this method can extend to any dataset registering the co-participation of actors and facilitate a comparative approach to the study of fields.