Global South Alliance Launches US$ 72,000 Fund to Support Research on Datafication and Democracy
Press Release – Tuesday, 09th December 2025: The Global South Alliance, a coalition of 26 digital rights organizations, launched today the second edition of the “Datafication and Democracy Fund” on December 9. The Fund will provide more US$ 72,000 to support research and advocacy projects focused on datafication and democracy to be implemented in 2026.
Transforming Classrooms by Leveraging Datafication for Effective Learning Analytics
In the evolving educational landscape, leveraging datafication for effective learning analytics is a game-changer. Educators and institutions now have the ability to turn vast streams of educational data into meaningful insights that enhance learner engagement, optimize outcomes, and refine instructional design.
Introduction to Datafication and Learning Analytics
Leveraging datafication for effective learning analytics involves transforming educational activities into quantifiable data. This empowers educators to track student progress, identify patterns, and intervene with precision. Rather than hoping for learning improvements, datafication enables strategic measurement and targeted support that align with core instructional objectives.
The Role of Datafication in Education
Datafication converts learning actions—such as resource usage and interaction timestamps—into measurable inputs. Through learning analytics, these inputs inform trends and identify obstacles. When educators ask how leveraging datafication for effective learning analytics works, they uncover ways to transform raw interactions into actionable intelligence, enhancing decision-making across pedagogical levels.
Driving Personalization with Learning Analytics
Incorporating datafication allows for highly individualized learning paths. Educators can analyze learners’ progress in real time to adjust pacing, recommend resources, and tailor feedback. This capability turns one-size-fits-all instruction into a flexible, learner-centered journey, where each student receives support tailored to their unique needs and pace.
Data-Driven Feedback Loops and Adaptive Learning
When datafication fuels learning analytics, feedback loops become dynamic. Analytics can highlight where a learner struggles and trigger adaptive content adjustments or timely intervention. Rather than static instruction, this approach nurtures continuous improvement and engagement through real-time responsiveness.
Improving Outcomes through Behavior Insights
Analyzing data patterns—like time spent on tasks, submission behaviors, or resource preferences—lets educators predict who may fall behind or succeed. Leveraging datafication for effective learning analytics enables early detection of disengagement and empowers educators to support learners with precision before challenges escalate.
Ethical Considerations in Learning Data
While datafication opens doors to deeper insight, ethics and privacy must remain central. Transparent data use policies, informed consent, and secure handling ensure that analytics support learners respectfully. Responsible practices safeguard trust while fully enabling the benefits of data-driven strategies.
Building Institutional Capacity for Analytics
To truly benefit from datafication in learning analytics, institutions need robust infrastructure. This includes training educators in data literacy, integrating analytics tools into workflows, and using data insights to drive policy and curricular improvements. Investments in capacity foster a sustainable, evidence-based learning ecosystem.
Leveraging datafication for effective learning analytics empowers education systems to evolve from intuition-driven practice to insight-driven strategy. By converting interactions into intelligence, educators can personalize learning, intervene precisely, and ethically steward student success. In this way, datafication becomes not just a tool but a partner in advancing learning.
The Political Erosion of Intelligence: Populism, Policy, and the Rewriting of Rationality – part 2
The persistence of political regimes founded upon the calculated disempowerment of critical intelligence is neither incidental nor anomalous; it constitutes, in fact, a necessity within the architecture of post-democratic populism.
If the first part of this post has sought to expose how populist logic systematically restructures public discourse, institutional credibility, and the very grammar…
Data increasingly forms the backbone of systems and processes that shape how we do things and how we relate to each other. Datafication – the uptake of data to reorganize social processes – is reshaping everything from loyalty programs and digital identification systems to credit card payments and rental pricing platforms. Artificial intelligence accelerates these processes.
Making sense of what these changes mean for our everyday lives is no easy task. Datafied systems are highly technical and designed to be convenient and seamless; we tend to encounter them in brief moments of individualized transaction, which makes them difficult to see, let alone read, and their illegibility makes them very challenging to respond to. Communing Data Literacy offers a novel set of concepts and tools to help people make sense of how technology is altering their communities and their social interactions. Building on three years of design research by digital rights organizations in Chile, Colombia, Paraguay, Peru, and Uruguay, the volume analyzes people’s everyday experiences with datafication, rethinking data from the perspective of community and offering practical techniques for community engagement.
Communing Data Literacy pushes back against the individualism and technocentrism of Western data literacy practice and scholarship, providing English readers the opportunity to engage with Latin American perspectives.
The year is 2025. For an average human living in the territories dressed with internet cables, the day starts by grabbing the smartphone and consuming whatever the algorithm has to offer. From grocery shopping to becoming a millionaire overnight through crypto trades, everything seems possible from behind the screen. Societies are increasingly shaped by the very algorithms that dictate behaviors, tastes, and desires. From the frenzy of aesthetic surgery trends to the instantaneous viral success of products, from the commodification of reality to the proliferation of memes, we have become subjects of a culture where everything is recontextualized, reshaped, and hyper-real. Our daily lives and social habits are shaped by the algorithm we constantly labor to. The lines between the real and the simulated blur further, as Baudrillard whispers from the early days of the internet, “We live in a world where there is more and more information, and less and less meaning.” (Baudrillard 1994:79). Here, meaning becomes a construct of virtuality, a mere image or simulation of the real. As our perception of reality becomes distorted in an AI-mediated fashion, whose pace of progress is beyond our perception of the pace of living, the human condition and social order are caught in a tension between the expectations of a world driven by accelerating technological advancements and the limitations of societies struggling to keep up. The contemporary condition whispers to us to either try and stay relevant or stay out of the picture. But even then, salvation is not guaranteed. In fact, nothing is guaranteed except the increasing quest for the relentless advance of an unchecked, accelerationist future.
This paper examines how the implications of contemporary accelerationist discourses of progress imply biopower and commodification of the subject by analyzing the longevity industry and public figures such as Bryan Johnson and viral self-optimization trends online. Through a critical analysis of the longevity industry, the paper aims to critically engage with the societal repercussions of accelerationist ideas.
The first book to draw a direct line between the datafication and prediction techniques of past eugenicists and today’s often violent and extractive “big data” regimes.Predatory Data illuminates the throughline between the nineteenth century’s anti-immigration and eugenics movements and our sprawling systems of techno-surveillance and algorithmic discrimination. With this book, Anita Say Chan offers a historical, globally multisited analysis of the relations of dispossession, misrecognition, and segregation expanded by dominant knowledge institutions in the Age of Big Data. While technological advancement has a tendency to feel inevitable, it always has a history, including efforts to chart a path for alternative futures and the important parallel story of defiant refusal and liberatory activism. Chan explores how more than a century ago, feminist, immigrant, and other minoritized actors refused dominant institutional research norms and worked to develop alternative data practices whose methods and traditions continue to reverberate through global justice-based data initiatives today. Looking to the past to shape our future, this book charts a path for an alternative historical consciousness grounded in the pursuit of global justice.
A free ebook version of this title is available through Luminos, University of California Press’s Open Access publishing program. Visit www.luminosoa.org to learn more.
Avete mai provato a chiedere a Chat-GPt-4: che fine fanno i nostri dati?
Siamo entrati nel web2 senza una grande consapevolezza degli utenti di internet in una infosfera in cui generiamo una quantità enorme di dati. Non solo quando navighiamo su Internet o usiamo i social, ma anche nelle azioni più normali della nostra vita quotidiana: quando ordiniamo un film su Netflix, quando utilizziamo la…
“It is easy to take for granted the value of data. It has come to seem self-evidently useful, as necessary and natural as water. It doesn’t even matter what has been measured and datafied; data in the abstract, as an idea, is taken to be a good thing, and of course there should be more of it, to enrich our knowledge of the world and to make anything that is “data-driven” work better. If data is being collected but not leveraged, why bother? Why have an archive of implosion images if not to simulate any implosion image imaginable?
But to accept that at face value would be to neglect the vast infrastructure involved not merely in collecting it and making it useful and tradable, but also establishing its reputation for objectivity. Measurement is an ideology; among its central tenets is that there is no such thing as datafication but just data itself, naturally given by the things in themselves. It presents itself as a form of representation that transcends representation: Data is no longer about the world but is instead taken to be the world itself, as though materiality were a matter not of atoms but of information. The image of an implosion is an implosion.
Likewise, this ideology would persuade us to ignore the market for data, which shapes what is measured and how, and have us believe it is more like a natural resource, a found material waiting for refinement rather than a structured informational good without any natural status at all. Implosions just happen.
Calls to measure everything and collect as much data as possible are offered as efficient strategies to better grasp the world as it is. But measurement is an act of power, not observation. Datafication always reifies an existing distribution of power that grants the measurers the ability to decide which aspects of the world count and which ones don’t. Having measurements taken as objective — having representations be treated as realities — requires power and recurrent processes of legitimation.”
…
Restating that in the terms outlined above, an archive recognizes the power relations intrinsic to measurement (and representation in general) whereas a dataset suppresses them (helping entrench the power relations that underwrite the data it assembles). An archive attempts to retain how and why representations were made, and a dataset disregards all that to allow representations to masquerade as universal facts. When representations become data, they reinforce the utility of the infrastructures (algorithmic decision-making systems, AI models, etc.) developed to exploit them. And that infrastructure in turn reinforces the power relations authorizing the data.
First they said they needed data about the children to find out what they’re learning. Then they said they needed data about the children to make sure they are learning. Then the children only learned what could be turned into data. Then the children became data 🙃
Datafication helps businesses unleash the true power of data and improves your business’ ability to predict strengths, weaknesses, potential, possibilities and outcomes accurately.
La Casa de Papel — Does it ring a bell? Are you familiar with this word? But it is surely in the top 5 of your favorite web series. What? Yes, you may not know the OG Spanish version, but you are a big fan of its English version, Money Heist!
The Spanish version was not a blockbuster. But Netflix translated the show not just into English but also into other languages. The first two seasons went on to become one of the most-watched web series without any promotion or advertisements.
It happened because of recommendation systems that have sophisticated algorithms with the proper tags and classification and user personalization, backed up by data science and machine learning. It is a classic example of datafication.
What is Datafication?
What is datafication — Is that even an acceptable English word? Before, it wasn’t, but it is today.
Datafication refers to the collective tools, technologies, and processes used to transform an organization into a data-driven enterprise. An organizational trend of defining the key to core business operations through a global reliance on data and its related infrastructure.
The crux is, “Datafication” is the process of turning everything into data. It is the act of taking something that was once unquantifiable and turning it into quantitative data.
Datafication enables the transformation of business operations, behaviors, and actions, in addition to those of its clients and consumers, into quantifiable, usable, and actionable data. This information can then be tracked, processed, monitored, analyzed, and utilized to improve an organization and the products and services it offers to customers. To put them into perspective.
Google transforms our searches into data
Facebook transforms our friendships into data
LinkedIn transforms our professional life into data
Netflix or Amazon Prime transforms our watched TV shows and films into data
Tinder transforms our dating activities into data
Amazon transforms our shopping into data
Data either personal or commercial are used to monitor every activity within its reach. Massive datasets are stored that get updated daily by the above tech giants for datafication. Collected data is then used for personalization in the form of ads, push notifications, consumable content, and more within each tech app or platform. This level of interference is usually regulated by the law.
The Datafication of Business
Data has now become a commodity. The currency is data. To produce it, tech companies bring together platform users who create data.
Datafication is a far broader activity, taking all aspects of life and turning them into data format. Once we datafy things, we can transform their purpose and turn the information into new forms of value — Big Data article (2013) by Mayer-Schoenberger and Cukier
Manufacturing and Supply chains
It simplifies the formation of short supply chains, creating micro supply chain business processes condensed through low-cost technologies such as mobile phones.
Real estate
It has made it possible for companies to gain in-depth insights into different locations, which in turn provides a better understanding to business leaders on where is the best place to locate their business.
FinOps
Managing financial activities across an organization is known as financial operations management (FinOps). Datafication is crucial because it enables the analysis and integration of data that was previously isolated in many systems. For example, datafication strives to bring together Accounts Receivable and Accounts Payable systems together to get a single view.
Human resources
Employers can identify potential employees and their unique traits, such as their risk-taking profiles and personalities, using mobile phones, apps, and social network data. Instead of depending on obsolete personality assessments or tests that gauge analytical thinking, it will replace existing exam providers.
Customer relationship management
Many businesses are using datafication to better understand their customers and develop applicable triggers based on their personalities and habits. This information is derived from the vocabulary and tone used in emails, phone calls, and social media.
AIOps
The phrase “AI-as-a-service” (AIOps) is used to describe how AI tools are employed in businesses. Another advanced technology that applies datafication to its domain is this one. Datafication combines a variety of AI tools and is cloud-based to deliver real-time data, insights, and measurements on nearly everything. You can use a web browser or a mobile device to access it.
Benefits of Datafication
Datafication offers enormous opportunities for improving business processes, making it a strategy that is financially advantageous to implement. Datafication is a new developing approach as well as a methodology for building a secure and innovative framework for the future of data analytics.
1. Actionable Insights
Datafication converts unstructured, incomprehensible data into usable insights, allowing you to get insight into your processes and procedures — the basis of any organization.
What do you do well? What needs to be improved? Conversely, what is working well but may be improved? Datafication implies that you will be more capable of understanding your company’s strengths, limitations, potential, and prospects. Also, it provides you with insight into the outcomes and ramifications of your projects, enabling you to assess what you’re doing and how you’re doing it.
2. Digital Transformation
Digital transformation services is no longer a fleeting fad; it is becoming increasingly crucial for all businesses that want to stay up-to-date and pertinent in an ever-changing ecosystem.
To take advantage of the latest and most cutting-edge technologies you should have usable data. It is the ticket to improving business processes and efficiency. It will help you to understand where the organization stands and the required next steps to move forward.
3. Improve Productivity and Efficiency
Datafication will comprehend what you’re doing and how you’re doing it better. Streamlining operations will make better use of all available resources, including employees, to boost overall production and efficiency and, as a result, transform your business into a successful enterprise.
4. Manage Information
Any business is generating a large amount of data and it is being collected and stored every day. If the data is managed well, it shall be providing better results. Otherwise, it can be overwhelming or can become unused data.
Datafication guarantees that you organize it appropriately, allowing you to properly use data to make decisions. You will not only be able to store data but also access and interpret it. Many businesses are experimenting with integrating user-sourced data and incorporating it into apps to contextualize the customer experience.
Conclusion
We know where you are. We know where you’ve been. We can more or less know what you’re thinking about — Erik Schmidt
The concept of datafication may be scary, but properly handled datasets with proper law regulations, security measures, and professional ethics could bring companies to provide customer-friendly and personalized services with the data collected. As datafication becomes more common it is driving innovation, breakthroughs, and betterment for the greater good.
One of the core elements to achieving datafication is by democratizing data access. Ensuring the last line of employees is empowered to access insights can build a data-driven culture that can act as a precursor for setting organizations on the path to datafication. Which brings us to the question, how does one democratize data access?
Is it possible? With the advancements that have happened around NLP, it’s very much possible. Listen to our webinar on how business intelligence can be reimagined using Conversational AI.
Data is everywhere, and datafication is transforming the way we live and work. From artificial intelligence to machine learning, discover how this powerful technology is unlocking valuable insights and driving innovation, while also posing new risks.
Datafication | The Future of Information Processing
Hello Readers! Today, we will talk about an interesting topic that’s been making waves in the tech world – Datafication. Simply put, Datafication refers to the process of turning everything, from everyday objects to human behaviors, into data that can be analyzed and used for various purposes. Read more
<p>Cop shit is seductive. It makes metrics transparent. It allows for the clear progress toward learning objectives. (“Badges” are cop shit, by the way.) It also subsumes education within a market logic. “Here,” cop shit says, “you will learn how to do this thing. We will know you learned it by the acquisition of this gold star. But in order for me to award you this gold star, I must parse you, sense you, track you, collect you, and—” here’s the key, “I will presume that you will attempt to flout me at every turn. We are both scamming each other, you and I, and I intend to win.” When a classroom becomes adversarial, of course, as cop shit presumes, then there must be a clear winner and loser. The student’s education then becomes not a victory for their own self-improvement or -enrichment, but rather that the teacher conquered the student’s presumed inherent laziness, shiftiness, etc. to instill some kernel of a lesson.</p>
<p>No wonder the traditional humanities classroom of “read things, come together and talk about them, and write papers about them” has disappeared in the age of cop shit. There’s no game to fix, no battle to win.</p>
This is a talk I recently had to give on datafication of culture (or actually on what cultural data analytics a’la Tallinn University could be like). Some of my colleagues thought they would like their students to read this, so posting the text here. Here is the video, and below is the transcript.
I propose discussing cultural data analytics via two broad questions, both of which I have filled with provocations that I hope will allow us to discuss - the implications and the politics of how we define concepts, - the power of those definitions shape the disciplinary and methodological space we operate in - and how that in turn suggests a positive and inclusive vision of cultural data analytics. I have my own answers to these provocations, but I am hoping that you will have yours, and that the CUDAN team, when assembled, will agree on the shared ones.
The two broad questions are seemingly simple:
How do we define cultural data analytics, given the extensive debates that have surrounded all of the words in this formulation? and
What is it that we want cultural data analytics to be do?
What is culture?
Ok, let’s start from the first big question, what do we mean when we say cultural data analytics. And to be systematic, we need to start with what do we mean, when we say culture.
Culture is according to Raymond Williams “one of the two or three most complicated words in the English language” (Williams 1983, 87). Other scholars are of much the same mind. Some even argue that the term is ‘so overused, that it is better to break it down into its component parts and speak of beliefs, ideas, artifacts, languages, symbols, art, or traditions.
In The Long Revolution, Raymond Williams offered three ways of defining culture: 1 the “ideal” definition, referring to the systems of valuation by means of which groups establish hierarchies, and subsequently judge the worth, of people, places, objects, institutions, and ideas; 2 the “documentary” definition, referring to the whole range of artifacts, both material and immaterial, produced by a group of people; 3 the “social” definition, referring to “a particular” or “whole way of life” i.e., the patterns of thought, conduct, and expression, prevalent among members of a collective.
Relying on the last one, which Williams appropriated from anthropology, John Fiske has argued that for cultural studies culture ‘is neither aesthetic nor humanist in emphasis, but political’. Politics in this case is the practice of living together, and we must be better at it, because, at the risk of sounding melodramatic - the alternative to living together is dying separately.
Methodological implications of how we define culture
Marek Tamm (2016) has suggested in his introduction to the book “How to study culture” that culture is not something that is passively available for researchers to come study it, rather it is constructed in the process of defining and making sense of it. Culture is thus created as an object of study and our definition depends on the disciplinary background of the researcher studying it.
A distinction that has had a strong impact on the study of culture is between culture as practice versus culture as a system of symbols and meanings. The first approach focuses on the processes of meaning making, and perhaps coincides with the definition of culture or cultures as particular ways of life. The second focuses on the more or less stabile forms and codes within the body of what can broadly be called “cultural texts”. Of course, ideally, we want to study culture as both – texts and practices. However, I think keeping this distinction in mind has analytical merit for the discussion at hand, because it highlights not only the methodological, but also the critical or the politico-economic implications that accompany both definitions. Let’s look at these
Critical implications of how we define culture
Culture as a way of life or as a set of everything created by everyone happens - to a disturbing extent - on corporately owned platforms, which are - post what we in my field call the API-apocalypse - closed rather than open for researchers, and make unreliable, difficult partners. They are also, as Jose van Dijck and TarletonGillespie (also this) have been saying for about a decade, not neutral intermediaries, but performative and constitutive infrastructures. Social media platforms, but also appstores shape the performance of social acts instead of merely facilitating them. This means that relying on data created and classified by these corporate platforms for making research inferences is quite problematic. Richard Rogers has called this an issue of vanity metrics. The data that corporations create reflects their needs and their version of a way of life, a culture, sociality. Their version is made of likes, follows etc, because those help measure impact and worth within the attention economy that social media has become. It is a partial rendering serving capitalist needs, wherein everyone is a laborer, a consumer or a commodity, often all three at once.
The version of culture as an assemblage of cultural texts, could be seen more as an issue of digitalizing heritage. This brings it its own can of worms, because it basically means participating in the datafication and metadatafication of culture, which as we’ll talk about shortly, is not necessarily a uniformly positive goal. Datafication, is usually conceptualized as the transformation of social action and many other previously unquantifiable aspects of the world into quantified data, which allows real-time tracking and predictive analysis (Mayer- Schoenberger & Cukier, 2013). Datafication, as we’ll shortly discuss in more detail, has a politics.
Basically, how we define culture implicates whether we want to use existing data or create data, which invites a rather different set of methods, and has a rather different set of risks, implications and ethics.
What is data?
Ok, this brings us to our second word in search of a definition. What is data?
Data is a concept that is most tightly linked to empiricism and positivism. We answer empirical questions by obtaining direct, observable information from the world. That direct observable information, often conceptualized as discrete units of information, is what is called data. Once we verify data, we get, from the positivist perspective - facts.
However, this only seems straightforward. Just like the definition of culture emerges out of and depends on the process of defining it, so is data made and not found. As Geoffry Bowker has famously said: “Raw data is both an oxymoron and a bad idea; to the contrary, data should be cooked with care (2005, p. 183-184). Lisa Gitelman and Virginia Jackson (2013) propose that the seductive power of the term raw data lies in it echoing a presumption that data come before fact, which suggests that data are the starting point for what we know, and that hence data must be transparent or objective.
Data as a thing, data as ideology
My friend and mentor prof. Annette Markham has argued (2016) that in academic discourse data operates on at least two levels – as a thing and as an ideology, both of which obscure the fact that data is not where meaning resides.
She argues that speaking of data as a thing is an ideological stance, which leads us to focus our attention on the wrong part of the process, we focus on what remains after we tidy, clean, condense and simplify and invites us to focus on pieces of text, or outcomes of interaction, distracting us from the point that this is not where meaning resides. Meaning, arguably, resides in the interaction not the outcome of the interaction, it resides in the making and consuming of the text, not in the text itself.
This doesn’t mean that data is useless, or we should not try to make data, it rather means that just as we need to be clear on what culture means for CUDAN, we need to be clear on how we cook data in this project. What tools do we make or use, and how do these tools function as frames or filters. Because tools carry the epistemic traditions they derive from (cf. this by Eef Masson, 2017) and most, if not all of the analytics tools used to study culture today were built by empiricists and positivists. To make this point clearer, I invite us to think about data through metaphors
Data metaphors
You have probably all seen and heard a version of “data is the new …”
Cornelius Puschmann and Jean Burgess suggest that there are metaphors of data as a natural force to be controlled (so here there are a lot of oil and water metaphors) and metaphors of data as a resource to be consumed (where they place food and fuel metaphors).
Metaphor scholars (Lakoff and Johnson 1980) have been saying for decades that metaphors function conceptually to not only reflect but to construct our experience of reality. If we say “data is the new oil” the comparison of terms builds or promotes a particular meaning. The term being defined (data) is connected to the supposedly more known term (oil). So if we think of petroleum oil then we think of it having to be drilled, which is dangerous to do, we think that world economies depend on it, that finding it unexpectedly will make you very rich, that you can make anything from it. If the comparison sticks, and everyone starts calling data the new oil, as they kind of have, it will work under the surface not only to reflect, but to influence how we think about data.
As Luke Stark andAnna Lauren Hoffman recently argued, the data metaphors, in particular the oil metaphor, invites specific data practices and specific approaches to data ethics. Liquid metaphors of data lakes, data oceans, data floods and data tsunamis tend to “forestall ethical or regulatory interventions by positioning data as uncontrollable” (Lupton 2013). But Stark and Hoffman also propose that we can look to common data metaphors to solve some of the regulatory and ethical problems we’re having with internet intermediaries abusing our data. If data is a natural resource, then perhaps we need to borrow from the ethical codes of forestry and think of data stewardship. If we think of personal data as of personal digital remains, maybe we need to borrow from morticians or doctors, and think of data care or data fiduciaries - fiduciary duty is the legal obligation of one party to act in the best interest of another.
Again, for the talk at hand, I want to ask – what kind of a data metaphor do we at Tallinn University want to operate with? Should we be satisfied with pre-existing metaphors, and live with what they illuminate and obscure about the world?
Methodological implications of how we define data
Something we hear repeated so often, is that the volume, velocity and variability of ‘big’ data has transformed how social research is conducted. More interestingly, it impacts what we think we are doing when we conduct said research. This, I think is the biggest methodological implication of how we define data. Do we think we’re cooking it? And what do we think it means if we’re cooking it? Some of my colleagues have noted a dangerous erosion of the role and meaning of interpretation in “data-driven” research (Markham 2016). If we agree that there is an erosion, and if we agree that reducing phenomena to data involves classification, which in turn obscures ambiguity and contradiction (Gitleman and Jackson, 2013), and if we think ambiguity and contradiction are important when speaking of and for cultures, then we need to think of how to bring them back. One option is to try to imagine an interpretivist data analytics.
Interpretivism rejects the view that meaning resides within the world independently of how people and groups interpret it. Typically, interpretivists advocate for context, which often means asking people things. This might not always be possible or wise with the types of projects we are imagining for CUDAN. However, the idea of context sensitivity or thick descriptions has been utilized in the more recent discussions on whether we can and should thicken our data.
Latzko-Tith, Bonneau and Millette (2017) say that thickening data means supplementing data with richly textured information, in other words, adding layers to them. Thick data is coated with several layers of rich metadata and paradata, so it is like an onion. Instead of points, thick data are whole little structured worlds. But we can think of thickening data also in terms of being more creative with what counts as data or what kinds of data we have, want, what we discard when we clean it, do we clean all of it, etc. My own experience working with data scraped from the Instagram API and Twitter API have highlighted this on a very personal level. Thickening or layering 90 000 image posts or 25 000 tweets with anything other than the metadata that the platform provides may seem impossible. But computational tools can also show you that the 90 000 images are from 180 accounts, or that in the 25 000 tweets include only 520 heterogenous ones that have been retweeted even once, which makes space for layering based on the computational power of the human brain. Basically, the argument is that layering embodies what interpretivism has learned from hermeneutics, the circular way of working a chunk of data and its context.
The critical (political-economic) implications of how we define data
Now, depending on how CUDAN decides to define data and go about cooking it will situate it at more or less problematic end of the spectrum of what can be called the political-economy of datafication. One of the best questions I heard two weeks ago at the AoIR conference in a methods session, was: “What evil things could be done with these new insights you have generated?” So, I think it is important that we too contemplate what evil things can be done with CUDAN, and what version of the datafication of culture and life we want to contribute to.
Many professionals and scholars see datafication as a revolutionary research opportunity to investigate human conduct. But, datafication is also heavily critiqued. A very poignant recent critique comes from Jathan Sadowski (2019), who recently published an elegant analysis of data as capital (as opposed to data as a commodity, which other work has done).
Sadowski argues that like social and cultural capital, data capital is convertible, in certain conditions, to economic capital. It adds new sources of value and new tools for accumulation. It also currently guarantees that those who already have a lot of this capital, like GAFA (Google, Apple, Facebook, Amazon) or BAT (Baidu, Alibaba, Tencent), will accumulate more, and those who don’t have it, are unlikely to amass any significant amounts of it.
Looking at data as capital allows him to notice that the data imperative, or the drive to accumulate all and any data from all sources, by all means possible, now propels how business is done and how governance is enacted. This means a total datafication of everything, by subjecting previously non-commodified and non-monetized parts of life to the logic of datafication and colonizing new spheres of life or new places in the world, so they can become sites of data extraction. So decisions like buying a company or launching a service are increasingly made for data potential, not because of revenue. Google gives primary school studenst free laptops or invests in healthcare or hosts all of Tallinn University’s emails and documents not because it cares, but because it is already or will very soon profit from all of that data. Extraction of data – and Sadowski is specific about calling it extraction and not collection or even mining, because calling it extraction highlights the exploitative nature of dataveillance, where data is taken without meaningful consent or fair compensation - is a core component of political economy in the 21st century.
What is cultural data?
Ok so this brings us to the end of the prompts and provocations around definitions, and implores us to ask what we mean when we say ‘cultural data’ and through addressing the methodological implications of defining cultural data – what we mean by cultural data analytics
That the computational processes of sorting and classifying people, places, objects and ideas have profoundly altered the way ‘culture’, as a category of experience, is practiced, experienced and understood, is something that many authors have addressed (Striphas 2015, Andrejevic, Hearn and Kennedy 2015). So, the question is - is there any other way to define cultural data than as the process and outcomes of the datafication of culture. And if there is none, then the question becomes, is there a way to shape how datafication of culture happens or to imagine alternative ways of datafying culture, because what we have now, is consolidation of the work of culture into the hands of a few powerful corporations, which, if we believe Ted Striphas, will lead to “the gradual abandonment of culture’s publicness”
Methodological and critical implications of how we define cultural data
If cultural data is the process and the outcomes of the datafication of culture, which is currently to a large extent governed by corporations for corporate interests, then this invites another question for CUDAN -
Do we need to come up with so called alt metrics for understanding culture? And what would those be?
I mentioned Richard Rogers (2018) work in the beginning of this talk. He proposes metrics that do not build on social media as a vanity space, but as one for social issue work. He calls them critical analytics. We can basically treat the past 15 years of social media as a case study for why we can’t rely on the metadata and the datafication models that corporations have created for their own needs, because analyzing those creates a particularly tilted view of the studied phenomena and makes CUDAN contribute to instead of subvert what is arguably currently wrong with the datafication of culture.
Epistemology of cultural data analytics
This definition work quite neatly introduces a bigger issue, which is what kinds of ontological, epistemological and axiological premises do we want cultural data analytics to have? We’ve talked earlier about bringing a certain interpretivist sensibility to data analytics, at least to our methods of cooking data, but I’m not sure we necessarily want to situate CUDAN fully in interpretivism. We also don’t want to situate it in what Christian Fuchs (2017) calls digital positivism, which he says does not connect “statistical and computational research results to a broader analysis of human meanings, interpretations, experiences, attitudes, moral values, ethical dilemmas, uses, contradictions and macro-sociological implications. And which he says means that it is just what Paul Lazarsfeld called administrative research predominantly concerned with how to make technologies and administration more efficient and effective.”
Instead, I would suggest, and Fuchs suggests, and frankly most authors who have studied social media for many years are suggesting a critical theoretical alternative. What does that mean?
What is it that we want to accomplish?
Ok this finally brings us to the second big question, which is, what do we want cultural data analytics to do? If we want to build critical cultural data analytics, then whatever else we want it to do, we will want it to challenge dominant assumptions and, ideally, change the world towards a better place. No pressure, right?
Looking across various academic, corporate and strategy documents big data analytics and cultural data is imagined to promise the following:
data analytics in general seems to promise to:
help us gain unprecedented insight into stuff –like public opinion, behavior patterns and relationships.
build a more ‘productive and intuitive’ user/consumer experience.
overall, there are a lot of vague but optimistic promises that we can do research that doesn’t exist yet, ask questions that do not exist yet, open up new avenues for inquiry
Within the realm of cultural analytics and digital humanities more broadly, the promises seem to be that we can:
digitally preserve and share cultural heritage. Which:
allows new discoveries that will transform our understanding of our cultures, identities, heritage and history.
make sure these cultures do not disappear;
make sure the heritage industry is relevant in the digital age
allow cultural differences and commonalities to be explored.
shed light on human history and the relationships between cultural and geographic areas.
Help us understand the dissemination of ideas and cultural phenomena and,
in relevant cases (such as in art fairs, universal exhibitions, or Olympic games), improve the management of current events.
introduce data-driven decision-making in the cultural sector (how to do this without adding to the accumulation of privilege and disadvantages, inequality, discrimination etc
provide arguments for the provision and allocation of public funding and measurement of its impact
Frankly, most of these do not sound like critical ambition. Some of these sound outright administrative, many descriptive, some interpretative.
So, again, the question for CUDAN is – which goals do we want to set for our version of cultural data analytics.
Do we want to say that cultural data analytics will help us understand culture better?
Does that mean that we think that the ways in which we understand it now are not good enough? And I am looking at Marek Tamm who has recently edited a whole volume on this. So, you know, provocatively I ask, what’s wrong with those ways of understanding culture? Did you know that training creating just one AI model for natural-language processing can emit as much as 600,000 pounds of carbon dioxide? (Strubell, Ganesh and McCallum 2019 via this). That’s about the same amount produced by 125 roundtrip flights between New York and Beijing. How can we make sure that cultural data analytics is better enough than the more eco-friendly alternatives to be worth it?
Do we want to say that cultural data analytics will be more efficient in understanding culture? That it will create more actionable insights both for researchers and for policy makers? That it will release us from the chains of stepping on the same rake and making the same mistakes?
That, in and of itself, is a great goal. Sheila Jasanoff has suggested that actionable data can problematize the taken-for-granted order of society by pointing to questions or imbalances that can be corrected or rectified, or simply better understood, through systematic compilations of occurrences, frequencies, distributions, or correlations. She speaks specifically of the power of the compilations of climate data, but surely this could be a great asset in the cultural sector as well.
Then again, here too, we can ask what that costs. Another example - AMS, Austria’s employment agency, is about to roll out a sorting algorithm built to increase efficiency. They ran statistical regressions to find out which factors were best at predicting an individual’s chances of finding a job. So they can stop giving support to those who are less likely to find a job. Like women and disabled people. The algorithm increases efficiency and offers highly actionable insights, as it ensures that the agency does not waste resources on giving support to people who will not, in the end, benefit from it. How can we make sure we don’t build this type of efficiency?
What should cultural data analytics be?
Ok so, lets reiterate. I presume that everyone’s answer to what cultural data analytics should be is different, and that is the point of asking these questions and raising these provocations, but let me clarify my take on it and offer some quick examples.
1. I think that while in abstract it makes sense to think of culture as both a practice and a set of texts, it is always also political in emphasis. I also think CUDAN would possibly benefit from a narrower definition of culture, or at least assigning different narrower definitions of culture to specific subprojects. What I’m trying to say is that it is not enough, and perhaps it is even a bad idea to try to combine what is usually called social analytics, i.e. analytics of the trace data cooked on and by GAFA (Google, Amazon, Facbook, Apple) or BAT (Baidu, Alibaba, Tencent) platforms and what is usually called digital humanities, i.e. analytics of digitalized cultural heritage data, and call it cultural data analytics. I don’t think that this is the innovation we’re looking for. My work in social media allows me to see the problematic aspects making inferences of platform data, but it also makes me weary at the ambition to turn cultural heritage data into platform ready data. I think combining these two will keep us stuck in the social media logic that has or will soon colonize all our data, so true innovation lies in coming up with alternatives. This is, of course, easier said than done. If we do want to engage with the “existing” data people generate on the platforms, then I do think that instead of using their data as evidence of practices or ways of life, we should critically analyze infrastructures.
Let me offer an example - Nic Carah and Dan Angus (2018) at University of Queensland are, working on a project that they call “critical simulations”. They engineer and scrutinize how Instagram’s algorithms process, classify and make judgements about cultural life. So they are trying to build the infrastructure to critically analyze it.
2. I think CUDAN needs to be adamant that it is cooking data, and careful in who elses cooking it consumes, as well as who it cooks for, and whom it cooks for for free (and this invites an open data discussion, which I didn’t have tome to go into, but we can in the Q & A). This means that we should set aside resources towards critical tinkering with existing tools, invention of new ones, a reimagination of metrics.
Let me offer another example. Trevor Paglen and Kate Crawford recently organized an artistic intervention called ImageNetRoulette (look here). Image Net is a huge database of photographs that is broadly used to train AI systems in how to recognize, categorize and classify. It is one of the more widely used training sets for machine reading. Among the 14 million images ImageNet was trained on, there were images of people that were sorted manually by humans like Amazon Turkers. They categorized what they saw based on their own biases, and their biases ended up in the algorithm. So while it is easy to imagine a cultural data analytics project that just uses an existing tool to generate some sort of a semi metaphorical rendering of what people represent on social media, ImageNetRoulette was conceived to expose the biases and politics behind the datasets and thus the AI that classifies humans. The project was hugely popular and made its point elegantly. People were labeled in racist, sexist, misogynist and otherwise judgmental terms. And it has already had an impact, the researchers behind ImageNet promised to delete more than half of the 1.2 million “people” images from the dataset.
3. I think CUDAN needs to be ambitious, but profoundly critical in setting goals, to avoid digital positivism, administrative research at the service of efficiency, as well as artsy vanity projects with limited social impact. I think it needs to commit to impact and data justice.
Pitting research questions or interests against each other is problematic, and I am not trying to suggest that everyone needs to study populism, climate change or alternatives to the particular version of capitalism we have, but maybe we should. At least in some way. Is there way to make a project about Estonian heritage cultures to be about the current debates surrounding Estonian forests. Is there a way to simulate and critique and then productively build alternatives to existing infratructures or data logics? In social media research datafication, appification and platformization have become almost curse words, yet in what I have read about the digitalization of cultural heritage, we seem to be hardly able to wait before everything is an app.
I feel like CUDAN has a decision to make. What kind of a project does it want to be. Critical? Descriptive? Computational? Administrative? I don’t think it can be all in equal measures. But I am very excited about the idea of a truly critical, contextual and ethical version of data analytics.
Does it exist? No. Can it be built? I believe so. Maybe this will be CUDANs gift to the world.
“How would you like your child’s education, personalised or standardised?” We The Educators is a project brought to you by Education International, The Alberta Teachers’ Association and The Canadian Teachers’ Federation– designed to start a new conversation about the future of public education. https://vimeo.com/216191791 Supported by a detailed literature review the project explores the impact…
