In this article, we analyse how the debate on the ‘refugee crisis’ has been constructed in Finnish news media and social media by using big data analytics. The study applies big data with the aim of exploring the dynamics between the mainstream news media and social media and the ways in which these dynamics shape and strategically amplify different understandings of the refugee crisis. The research highlights over-emphasis of crime and threat-oriented themes on refugee issues in social media, as well as illuminates the distinct role of social media platforms in shaping debates through user practices of hyperlink sharing and networked framing. Together these findings suggest that the hybrid media environment provides a possible arena for polarization of the refugee debate that could also be used for political ends.

Scholarship on algorithms has drawn on the analogy between algorithmic systems and bureaucracies to diagnose shortcomings in algorithmic decision-making. We extend the analogy further by drawing on Michel Crozier's theory of bureaucratic organizations to analyze the relationship between algorithmic and human decision-making power. We present algorithms as analogous to impartial bureaucratic rules for controlling action, and argue that discretionary decision-making power in algorithmic systems accumulates at locations where uncertainty about the operation of algorithms persists. This key point of our essay connects with Alkhatib and Bernstein's theory of 'street-level algorithms', and highlights that the role of human discretion in algorithmic systems is to accommodate uncertain situations which inflexible algorithms cannot handle. We conclude by discussing how the analysis and design of algorithmic systems could seek to identify and cultivate important sources of uncertainty, to enable the human discretionary work that enhances systemic resilience in the face of algorithmic errors.

This paper investigates how unsupervised machine learning methods might make hermeneutic interpretive text analysis more objective in the social sciences. Through a close examination of the uses of topic modeling—a popular unsupervised approach in the social sciences—it argues that the primary way in which unsupervised learning supports interpretation is by allowing interpreters to discover unanticipated information in larger and more diverse corpora and by improving the transparency of the interpretive process. This view highlights that unsupervised modeling does not eliminate the researchers’ judgments from the process of producing evidence for social scientific theories. The paper shows this by distinguishing between two prevalent attitudes toward topic modeling, i.e., topic realism and topic instrumentalism. Under neither can modeling provide social scientific evidence without the researchers’ interpretive engagement with the original text materials. Thus the unsupervised text analysis cannot improve the objectivity of interpretation by alleviating the problem of underdetermination in interpretive debate. The paper argues that the sense in which unsupervised methods can improve objectivity is by providing researchers with the resources to justify to others that their interpretations are correct. This kind of objectivity seeks to reduce suspicions in collective debate that interpretations are the products of arbitrary processes influenced by the researchers’ idiosyncratic decisions or starting points. The paper discusses this view in relation to alternative approaches to formalizing interpretation and identifies several limitations on what unsupervised learning can be expected to achieve in terms of supporting interpretive work.