Skarpelos, Y., Messini, S. (2026). Cultural and Emotional Grounds for Image Choice in Twitter: The Case of #omicron Variant. In: Karpouzis, K., Skarpelos, Y. (eds) Computational Methods for the Digital Humanities. Learning and Analytics in Intelligent Systems, vol 57. Springer, Cham. https://doi.org/10.1007/978-3-032-05099-1_1
Computational methods challenge conventional visual analysis tools, inviting to a deeper understanding of uses of visual images in massive volumes by audiences themselves. Audiences have always used found images or amateur photographs, but such practices intensified with the advent of Web 2.0 and social media. More and more they are understood as a kind of self-motivated, amateur curatorial work. In this paper we explore the cultural roots and emotional motivation behind image selection by mass audiences in social media. We take an instance in the COVID-19 pandemic as a case study: images used in Twitter the day Omicron variant was announced. Using computational methods, we collected and analysed approx.23,000 images. We found a divide between textual and pictorial images, which may be explained based on cultural grounds. We also found that different kinds of imagery were related to tweets expressing differences in certain emotions like Fear or Sadness, and on secondary emotions like Despair, Guilt and Aggressiveness. The findings show that the connection between the visual, the textual, and the emotional elements is a fruitful field for novel knowledge.
Karpouzis, K. (2026). AI, Digital Humanities, and the Legacies of Colonial Power. In: Karpouzis, K., Skarpelos, Y. (eds) Computational Methods for the Digital Humanities. Learning and Analytics in Intelligent Systems, vol 57. Springer, Cham. https://doi.org/10.1007/978-3-032-05099-1_4
This paper examines the impact of artificial intelligence (AI) on Digital Humanities through a colonial lens, analyzing how AI can both reinforce and challenge colonial power dynamics. AI tools in Digital Humanities, such as text mining and language preservation, often perpetuate Western epistemologies and marginalize non-Western perspectives due to biases in data and algorithms. Using case studies, such as the Slave Voyages database and indigenous language preservation projects, this paper highlights AI’s dual role as both a potential perpetuator of colonial legacies and a tool for decolonization. It recommends inclusive AI design, community-driven data governance, and the integration of alternative epistemologies to mitigate AI’s colonial biases and promote more equitable knowledge production.
Skarpelos, Y., & Messini, S. (2026). Of Algorithms, Emotions, Images, and Semiotics: The Case of the COVID-19 Pandemic. Proceedings of the 15th World Congress of the International Association for Semiotic Studies (IASS/AIS), Thessaloniki 2022, 30 August – 4 September, University of Macedonia, Greece. https://iass-ais.org/wp-content/uploads/2026/05/IASS-2022-Congres-Proceedings.pdf
In this paper we examine the relationship between the characteristics of the images and the emotional content of the text accompanying them. The underlying research question could be formulated as follows: “Are images with similar figurative or plastic features related to texts expressing similar ideas or sentiments?” To answer this question, we first provide a “distant viewing” explanation. We then turn to the texts to corelate sentiments, emotions, and feelings to the images themselves.
Voulgari, T., Sdralli, I.E.S., & Skarpelos, Y. (2025). Digital Fashion and Social Identity. Contemporary Issues in Digital Communication, Media, and Culture, pp 223-263. https://doi.org/10.4018/979-8-3693-8729-0.ch008
The global fashion system changes decisively due to the impact of the growing technological innovation and social changes. Fashion is a method of self-expression that gives people the opportunity to be their true selves.The purpose of this study is to present a new perspective about the combination of fashion and gaming. The connection between the two seemingly different “worlds” is more relevant than ever considering that over three billion people worldwide are actively participate in-game experience. Further research seems necessary to explore how virtual fashion is used for new forms of self-identity. To develop a research model for approaching this aim, the study employs four theoretical approaches. So, this systematic literature review investigates the evolution of avatar fashion, the innovative ways that virtual environment used for forms of social identity, the collaboration between fashion-gaming industries and the Metaverse as a digital universe of avatars.
Karpouzis, K., & Tsatiris, G. A. (2022). AI in (and for) Games. Advances in Machine Learning/Deep Learning-based Technologies: Selected Papers in Honour of Professor Nikolaos G. Bourbakis–Vol. 2, 27-43. https://doi.org/10.1007/978-3-030-76794-5_3
This chapter outlines the relation between artificial intelligence (AI)/machine learning (ML) algorithms and digital games. This relation is two-fold: on one hand, AI/ML researchers can generate large, in-the-wild datasets of human affective activity, player behaviour (i.e. actions within the game world), commercial behaviour, interaction with graphical user interface elements or messaging with other players, while games can utilise intelligent algorithms to automate testing of game levels, generate content, develop intelligent and responsive non-player characters (NPCs) or predict and respond to player behaviour across a wide variety of player cultures. In this work, we discuss some of the most common and widely accepted uses of AI/ML in games and how intelligent systems can benefit from those, elaborating on estimating player experience based on expressivity and performance, and on generating proper and interesting content for a language learning game.
Korompili, A., & Karpouzis, K. (2022). An Early Childhood Introduction to Robotics as a Means to Motivate Girls to Stay With STEM Disciplines. In Handbook of Research on Integrating ICTs in STEAM Education (pp. 22-40). IGI Global. DOI: 10.4018/978-1-6684-3861-9.ch002
This research examines the design, implementation, and impact of an educational robotics intervention for first and second grade students. It controls for gender-related performance differences and compares the interest shown towards robotics. The authors also examine if factors such as students’ stance towards different professions can contribute to a difference in performance. In the course of its work, custom designed worksheets for the UARO educational robotics product were used, as well as questionnaires given to students after meetings. The results showed that all genders responded equally well and with the same enthusiasm to the robotics activities and understood concepts of physics, mechanics, and mathematics through them. Participants differ in how they use their leisure time and in their professional orientation; however, this didn’t affect their performance in the robotics activities. These results highlight the need for further examination of the social institutions and factors that influence the formation of gender orientations during the early childhood age.
Roinioti, E., & Pandia, E. (2019). Planet Dewey: Designing a Hybrid Game to Boost Students’ Information Literacy. In Games and Learning Alliance: 8th International Conference, GALA 2019, Athens, Greece, November 27–29, 2019, Proceedings, Springer International Publishing. DOI: 10.1007/978-3-030-34350-7_6
This paper presents a case study of the key design choices that were made during the development of «Planet Dewey», a 2D hybrid co-op game for students aged 11–14 years old for strengthen their information literacy skills. The game was tailor made for a Greek charitable foundation library. Using as a starting point the information literacy circle and through puzzle-based missions, scavenger hunt elements and an age-appropriate narrative structure, we aimed to increase intrinsic and extrinsic motivation. Experimental design through controlled frustration was another aspect of our approach.- Skarpelos, Y. (2018). Big Visual Data in Social Sciences. In Welker, M., Stützer, C. M., & Egger, M. (Eds.). (2018). Computational social science in the age of big data: concepts, methodologies, tools, and applications. Herbert von Halem Verlag.
Visual social sciences have considered themselves as part of the qualita tive tradition and have therefore dealt with >small< visual data. Extensive visual data digitization programs and the explosion of social media gave rise toa miner quantitative tradition and offered the excitement of dealing with >big< visual corpora. In this chapter I first try to answer the question when are visual data big?< based on examples of current research. Then I provide an account of technologies appropriate to visual research and state-of-the-art publications using such technologies. Data visualization is an impoJtant part of big data literature. It is more soin big visual data research, and I present examples in the use of such visualization tech niques as composite images, image montage, histograms, and entourage plots. The last section is dedicated to the discussion about the use of such technologies and techniques in the framework of traditional visual social sciences' methods like content analysis and semiotics.- Roinioti, E., Pandia, E., & Skarpelos, Y. (2017). Sociability by design in an alternate reality game: the case of the trail. In G. Niemeyer, A. Garcia (eds.), Alternate Reality Games and the Cusp of Digital Gameplay, Bloomsbury.
Alternate Reality Games (ARGs) challenge what players understand as “real.” Alternate Reality Games and the Cusp of Digital Gameplay is the first collection to explore and define the possibilities of ARGs. Though prominent examples have existed for more than two decades, only recently have ARGs come to the prominence as a unique and highly visible digital game genre. Adopting many of the same strategies as online video games, ARGs blur the distinction between real and fictional.With ARGs continuing to be an important and blurred space between digital and physical gameplay, this volume offers clear analysis of game design, implementation, and ramifications for game studies. Divided into three distinct sections, the contributions include first hand accounts by leading ARG creators, scholarly analysis of the meaning behind ARGs, and explorations of how ARGs are extending digital tools for analysis. By balancing the voices of designers, players, and researchers, this collection highlights how the Alternate Reality Game genre is transforming the ways we play and interact today.- Skarpelos, Y. (2017). Towards a Quantitative Visual Semiotics? In K. Bankov et al. (eds.), New Semiotics Between Tradition and Innovation: Proceedings of the 12th World Congress of the International Association for Semiotic Studies (IASS/AIS), IASS Publications & NBU Publishing House. DOI: 10.24308/iass-2014-042
Visual Semiotics, as is the case with most of Visual Studies so far, have not found yet a sound and methodologically rigorous equilibrium between ‘visual’ and ‘semiotics’. Instead, following existing examples, they scrutinize visuality, the gaze and other similar objects of inquiry, or work on visual grammar and syntax. In all cases, they deal mostly with meta-language about images than with images themselves. They deal mostly with words used to describe the visual (either produced by the researcher or by others) than with the visual itself. In this paper, I am to propose the use of existing software tools in order to achieve research in quantifiable signifiers in visual images, and their manipulation towards a quantitative semiotics. Concepts like ‘entropy’, ‘hue’, ‘saturation’, ‘mean RGB’ etc. are the computed signifiers which are actually visible but so far visual researchers were unable to tap upon. Illustration of the method will be provided by my current research.