Algorithmic Life: Calculative Devices in the Age of Big Data by

Language: eng
Format: azw3
ISBN: 9781317527374
Publisher: Taylor and Francis
Published: 2015-12-21T16:00:00+00:00

The first step in our project was to measure the trackers on a dataset of digital subprime websites at two separate points in time (July and November 2013) and then to simply count the ‘kinds of data’ that we encountered. To do so, we employed Ghostery’s categorisation, as outlined above. The results (Figure 5.2) reveal that, for these seven websites, trackers involved in the collection of browser information featured most prominently (60 times) as a collected data type. We already knew from our initial experiments with Wonga that different browser types were affecting the slider position. Here we see an interest in browser information across the sector. Other prominent data types include the date and time of a visit, as well as the users’ IP address and hardware/software type. From conversations with those working in the industry we know that these are of potential interest to those working within digital subprime; here we can begin to render this interest visible.

In this short pilot experiment, however, we have been left with as many questions as answers: what accounts for the broad increase in trackers being used across the sector? Is this a general upward trend, or an anomaly? How do Ghostery’s categories translate into what data is being actually connected. For instance, the results show trackers being involved in the collection of phone numbers (e.g., from mobile devices) – however, this may well relate to particular online telephony services, or to when phone numbers are volunteered by users.19 Similarly, any demographic data being collected may not necessarily be able to be tied down to the level of the individual.20 Moreover, there references to PII (Personally Identifiable Information) connect the categories to a particular term within US legal discourse – in Europe the preferred category is ‘personal data’ and does not necessarily refer to the same kinds of data (Borgesius, 2013). To a degree, then, these results are a prompt for further research.

That said, there are further ways in which this data can be set to work. By tracker tracking, we have been able to begin to develop profiles of both the work that individual trackers are undertaking, as well as the particular forms of tracking each website may be mobilising by combining individual trackers. For each is involved in a quite distinct form of “socio-technical knitting”, drawing on Jose Ossandon’s term (Ossandon, 8 July 2013). That is, they are pulling together different online strands to each compose unique invisible ‘tracking fabrics’. The question, however, is how to render this quite abstract, technical work visible and communicable.

For this, we can also turn to Ossandon. In some research on the ways that credit cards are passed between individuals and households in Chile, Ossandon (2012) asked his participants to map the whereabouts of their retail cards by pinning down woollen threads. Here some of the knitted socio-economic relations surrounding credit in Chile become visible and different profiles of movement comparable. In our case, the socio-economic knitting is undertaken not just by people but also by quite specific combinations of trackers.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.