Digital Data Infrastructures: interrogating the social media data pipeline

Abstract

Far from being 'naturally occurring', as is sometimes claimed, this paper begins from the premise that social media data are shaped through complex social and technical processes: from creation, to access, storage, analysis and visualization. Drawing on the vocabulary of Science and Technology Studies we conceptualize this as a sociotechnical infrastructure and offer a methodological description of this using the metaphor of a 'data pipeline'. Using this framework, our paper draws on a combination of evidence from secondary literature and our own experiments, our paper aims (i) to demonstrate the processes at play at each stage of the pipeline (ii) to interrogate the iterative nature of the pipeline (how decisions made at one stage shape research possibilities at a later stage) and (iii) to assess the consequences for social research and the value of social media data in this endeavor.