Authors:
Table Coordinators:
Margarita Gutman and Nicolás Quiroga
Exhibitors:
Emmanuel Borthiry, Emiliano Calomarde, Javier Nesprias, Marcelo A. Soria, Margarita Gutman, Martin Pablo Groisman, Milagros Dolabani, Nicolás Quiroga, Silvana Ferreyra, Wanda Juares and Ximena Puppo.
Link:
https://youtu.be/j8nP057sUno
Summary:
Mateo Pasquinelli described the (future) society as the metadata society, understanding this term as a device of control and subjectivity. The simplicity of the vulgate term metadata (“data on data”) retains social problems inscribed in long-term structural technical processes and in “digital” processes that, since the nineties of the 20th century, have amplified the struggles to make the information classification systems meaningful and “natural”.
At present, the opacity of complex algorithmic systems that work with large amounts of data through AI / deep learning and the naturalization of the circulation and consumption of data produced by large corporations of platform capitalism demand a critical look at metadata. The same requirement applies to the “black box” of programming and to the use of “shielded” algorithms in the analysis of information.
Exploring and reflecting on metadata opens up the possibility of denaturing an uninformed understanding of situations that are presented to us as objective, universally accepted or definitive.
One example, among many, is the one presented by the images of Google Street View. These are intended to represent the visualization of a tour through the streets of the city at a given moment, which we assume is the one that appears in the shot that we observe. However, a more detailed search shows that this visualization, although it corresponds to the streets in question, is made up of shots taken at different times. In other words, the images that are presented as a synchronous visualization of the state of some streets, actually constitute an asynchronous composition. In this way, the understanding of the temporality of the representation is altered, which, although it may be irrelevant in some cases, in others may be decisive for decision-making or for the course taken by an analysis process.
The analysis of the metadata, which could be considered as a retracement of the data construction process, is crucial for the understanding of the production, circulation and pregnancy in social life of the algorithms that organize information in our present.
Convinced of this imperative, the table proposes to reflect on current research and future projects on methodological and political problems and dilemmas in research on data construction.
Presentations:
A look from the forges: building data, metadata, and annotations to train machine learning algorithms. The renewed potential of metadata
Marcelo A. Soria
(FAUBA, FCEN, UBA)
“PSYCHO / DATA / GEO / GRAPHICS”. Public space as a psychogeography of private data
Martin Pablo Groisman
(Expressive Media – Graphic Design, FADU – UBA)
(Self) learning processes of digital tools in social research: word clouds on Twitter
Emmanuel Borthiry, Emiliano Calomarde, Milagros Dolabani, Silvana Ferreyra and Wanda Juares
(INHUS, Conicet)
A messy world. Archiving, reading and algorithms in working with digital repositories
Nicolas Quiroga
(INHUS, Conicet)
Territories and temporalities of the digital city. Exercises to denature Google Street View and its data
Margarita Gutman, Javier Nesprias and Ximena Puppo
(Latin American Observatory, The New School – IEHu FADU – UBA)