Convenors: Biagio Aragona & Adam Arvidsson (University Federico II of Naples)
As algorithms and (big)data are increasingly shaping government policy, commercial practice and popular representations of the social world, and the need to think critically about their production and consumption is ever more recognized.
In the past ten years a growing number of scholars have started to focus critical attention on software code and algorithms contributing to Science and Technology Studies (STS), new media studies and software studies, in order to unpack the nature of algorithms. Some have concentrated on how algorithms are generated (Bucher 2012; Geiger 2014), others on how they work within specific domains such as journalism (Anderson 2011), security (Amoore 2006; 2009) or finance (Pasquale 2015).
An empirical effort to investigate the intertwined assemblage of socio-cultural and material entities that shape algorithms and data can be fruitfully detected in ethnography. Within STS (Latour 1987), similar methods have been used for a long time to define data infrastructures (Star and Ruhleder 1996). They are influential in more recent work on the “social life of data” (Ruppert 2015), or “data assemblages” (Kitchin and Lauriault 2014). Ethnography has also been employed in critical data studies (Iliadis and Russo 2017), which aim to interrogate all forms of potentially depoliticized data science, in order to track the ways in which data are generated, curated, and how they permeate and exert power.
An interesting example of this algorithmic “power” is the case of teacher mobility in Italy, where the ministry of education applied a not open source algorithm in order to reallocate school teachers on the national territory according to school needs. An aspect of the process was particularly disapproved by the teachers: the not open source nature of the algorithm, and the difficulty to verify it. As reactions, lots of appeals in court were advanced in order to proceed towards accessing the algorithm as a computer translation of an administrative procedure. This case shows that the data assemblage results from a complex network of relationships where all the actors involved (human and non human) play a specific – and sometime an unexpected role. Understanding the role played by all these actors is one of the possible object of ethnographic research on data and algorithms.
Ethnography seems suit to face the challenges that are head. The intensive use of massive databases and the wide application of algorithms have risen some concerns, because they may lead to a technocratic form of governance (Mattern 2013). A further risk is that big data may accelerate a process of corporatization of the public arena, because they are mainly private data coming from the largest software and hardware services companies and from the big majors of communication and logistic. Unpacking the data assemblages and the performativity of algorithms in decision making may be one way for increasing the system responsiveness instead of reproducing new forms of technocratic regimes.
Our session welcomes ethnographic research addressing following, non-exhaustive, list of topics:
- Algorithms at work;
- Auditing algorithms
- Building algorithms;
- Construction of meaning from automated Natural Language Processing;
- Retracing cultural, symbolic and normative values in data and algorithms;
- Data Coding;
- The dynamics of data teams
- Definition of commands for Artificial Intelligence;
- Design of machine learning algorithms;
- Digital platforms’ analytics design;
- Ethics in automated data production;
- Practices of (big)data management and (big)data curation;
- Practices of access to data and algorithms;
- Pre-analytics of big corpora;
- Social networks as source of data;
- Software Coding.
The role of ethnography on such topics has been recognized (Geiger 2017, Seaver, 2017), and interdisciplinary research in sociology, anthropology, computer science, STS, philosophy and technology, human geography, digital humanities, and data science has been carried out. We welcome therefore scholars from all these disciplines.
- What are the practices of algorithms and data production and use?
- What are the actors who participate in the design of algorithms, and how do they interact?
- What values are incorporated in algorithms and data?
- How do communities of experts and stakeholders impact the shaping of algorithms and data?
- How may technical issues impact the shaping of algorithms and data?
Algorithms, (Big) data, Coding, Machine Learning, Data curation, Natural Language Processing, Artificial intelligence.
Fields of Study
Sociology of algorithms, Critical data studies, Technoscience, Digital ethnography, Social Data science.