Affective computing scholarship and the rise of China: a view from 25 years of bibliometric data

Abstract

Affective computing, also known as emotional artificial intelligence (AI), is an emerging and cutting-edge field of AI research. It draws on computer science, engineering, psychology, physiology, and neuroscience to computationally model, track, and classify human emotions and affective states. While the US once dominated the field in terms of research and citation from 1995–2015, China is now emerging as a global contender in research output, claiming second place for the most cited country from 2016–2020. This article maps the rhizomatic growth and development of scientific publications devoted to emotion-sensing AI technologies. It employs a bibliometric analysis that identifies major national contributors and international alliances in the field over the past 25 years. Contrary to the ongoing political rhetoric of a new Cold War, we argue that there are in fact vibrant AI research alliances and ongoing collaborations between the West and China, especially with the US, despite competing interests and ethical concerns. Our observations of historical data indicate two major collaborative networks: the “US/Asia-Pacific cluster” consisting of the US, China, Singapore, Japan and the “European” cluster of Germany, the UK, and the Netherlands. Our analysis also uncovers a major shift in the focus of affective computing research away from diagnosis and detection of mental illnesses to more commercially viable applications in smart city design. The discussion notes the state-of-the-art techniques such as the ensemble method of symbolic and sub-symbolic AI as well as the absence of Russia in the list of top countries for scientific output.

Reading joy, the Bruce Lee way

“…The Ph.D. candidate has another reason, as he recently had several of his important research papers published in prestigious journals such as Humanities and Social Sciences Communications (Nature Portfolio) [2], AI & Society [3][4], and Technology in Society (Elsevier) [5].…”

The Mindsponge-based extensions of the Technology Acceptance Model: Behavioral determinants of the attitude toward of non-conscious emotional data harvesting

“…Second, is the differing views of different business segments such as marketing where the use of emotional data is primarily considered a tool to improve customers' satisfaction (McStay, 2020). Third, is the cross-cultural incongruencies in interpreting the impacts of smart technologies on values such as autonomy and privacy (Ho et al, 2021;Mantello et al, 2021). And finally, the shaky science behind these affect-sensing algorithms which are trained with non-representative data raises questions of accuracy and relanvance for mutiple marginalized groups such as women, ethnic minorities, people of colors, and people with disabilities, etc (Barrett, 2021;Birhane, 2020;Chen et al, 2018;Crawford, 2021).…”

Sociocultural factors influence young people’s attitude towards the harvesting of non-conscious emotional data

“…Emotional AI, a term popularized by McStay (2018) and the Emotional AI Lab (2020) to characterize technologies capable of sensing, perceiving, learning, and interacting with human emotions and mental dispositions, has its roots in the 1990s when Rosalind Picard (1995) founded the field of affective computing (Ho et al, 2021). The new technology would collect a wide range of "emotional" data, moving beyond conventional words and images and toward subtle facial expressions, gestures, voices, and body movements.…”