The concept of "big data" was introduced in the 1990s. Since then, tech companies, financial institutions and governments have been working with massive, diverse data sets to uncover patterns, extract insights and predict outcomes. Decades of digitisation and the rapid rise of social media content, have given AI companies the massive datasets they need to power their technologies.
Visiting Professor in Moa Martinson’s name at LiU, Theopisti Stylianou-Lambert, and Professor Bodil Axelsson explains why different forms of data have become such an important topic right now:
“Freely available data are not necessarily quality data, and there is a growing recognition of this. At the same time, the technology industry remains hungry for more data to consume, and "closed data seems to be the next target.”
Examples of “closed data” include undigitized materials, data behind paywalls, or otherwise inaccessible material.
“At the same time, one of the key concepts gaining traction in the technology industry is "small data": high-quality, targeted information that answers critical, high-impact questions.”
Researchers contribute to ethical and responsible AI
Researchers from the arts and humanities have traditionally worked with small- and medium-sized datasets, often using qualitative methodologies that yield rich data.
“Applying critical studies, feminist and post-colonial theory, research has shown how technological systems tend to reinforce past power imbalances. These studies also draw attention to issues of exploitation, data ownership and ethics.”
“This questioning of the uncritical use of big data gave birth to critical data studies that draw heavily on critical theory to address power structures and explore the social, cultural, and ethical challenges when working with big data.”
Researchers in the humanities, as well as technology experts, have identified clear limitations in big data approaches.
“We need to ask: How can researchers from the arts and humanities – who have an intimate knowledge of how photographic images/archives work, and the context and social practices of photography – use their expertise with small, rich data to (re)shape AI systems? How can we help create a more ethical, responsible and useful AI?”
