With the digitalisation of society, attention becomes mediated by, and deployed within, an ecosystem of devices and practices which enables traceability thereof. Drawing on this, we will present and discuss an ongoing research we carry out at the STSLab in Lausanne, which studies collective attention regimes by repurposing Wikipedia consultation data. Specifically, we built a database comprising ten years of all pages’ consultations from 41 language editions. We have also developed dedicated software to visually explore these data across space and time and identify, characterise, and compare with great precision the patterns of attention (consultations over time) elicited by very different kinds of entities (articles) and their relative ‘globality’ or cultural/linguistic entrenchment (language editions and geo-located content). On this basis, we analyse and discuss kinds of ‘matters of attention’ and their patterns. Thereby, we further research in STS which used Wikipedia page edit data for controversy mapping. Yet, by looking at attention patterns in and through the whole of Wikipedia – and not to edits in pages we know controversial–, we foster a non-aprioristic approach to what (i.e. articles) constitute matters of collective attention – and sometimes concern. More generally, we further epistemological reflexions on repurposing which are central to STS engagements with big data. Considering Wikipedia both as a site and a proxy, to what extent can the attention to digital media (as information queries) and the social phenomena of attention can be disentangled or assimilated?