TeamMeUP: Assessing the potential of computational matchmaking for the game development sector 

Principal Investigator

TeamMeUp, funded by The Dutch Research Council,  explores the potential of computational solution for team formation in the video game industry. Video game development is a creative sector with significant economic impact, relying on the efficient collaboration of creative professionals from diverse backgrounds to develop the next successful game. However, assembling an efficient team of these professionals is a challenge, leading to considerable economic and social consequences for the industry. Using optimization algorithms and insights from social sciences, TeamMeUp aims to help address the issue of inefficient talent coordination in the game industry. 

Timeline: 2025
Partners: Utrecht University
Keywords: team formation optimization, computational matchmaking, talent coordination, game development industry 

Teacher Support for Integrating Generative AI practices in Programming Courses

Co-principal Investigator

Building on the 2023-2024 USO project “Generative Artificial Intelligence in Programming Education”, this project focuses on integrating generative AI into programming education, with an emphasis on teacher training. Through workshops and empirical validation, it aims to align AI practices with course objectives and support evidence-based teaching in higher education. The end result will be a methodology that helps teachers align their use of AI in programming education (including code explanation, exemplar code generation, review of student code, exercise quick start, code refactoring, and analogies generation) with the intended learning outcomes, and the didactical approaches they wish to follow. 

Timeline: 2024 - present
Partners: Utrecht University
Keywords: programming education, teacher training, evidence-based teaching, generative Artificial Intelligence

GameDevUp: Upgrade Team Formation in Game Development

Supervisor

In the game development sector, the formation of effective teams has long been a challenge, often hindering creative potential and leading to inefficiencies. With the support of the Marie Skłodowska-Curie Actions, the GameDevUp project tackles this issue by exploring collaboration dynamics and developing an algorithmic matchmaking system to reshape how game developers connect and innovate. The project aims to provide a methodology that helps developers and incubators foster diverse, productive teams across geographic, social, and economic boundaries, while contributing to the broader literature on what drives successful collaboration in creative industries. Its ultimate goal is to deliver a scalable, algorithm-mediated matchmaking model that can enhance collaboration during game jams and incubation programs, with potential to evolve into a startup.

Timeline: 2024 - present
Partners: Utrecht University, Institute of Digital Games-University of Malta
Keywords: collaborative and social computing, computer-supported cooperative work, video games industry, creative industries

Generative Artificial Intelligence in Programming Education

Co-principal Investigator

​The goal of the "Generative AI in Programming Education" project is to explore the implications of generative AI tools, with emphasis on Copilot and ChatGPT, on programming education and to develop a web toolkit with resources and best practices to help teachers understand and successfully embed these tools into their teaching and assessment practices. The project is supported by the Utrecht University's Education Incentive Fund (Utrechts Stimuleringsfonds Onderwijs USO).

Timeline: 2023 - 2024
Partners: Utrecht University
Keywords: AI-enhanced software development, generative artificial intelligence, education transformation, future-proof programming

Crowd-AI Teams for Engineering Product Design

Co-principal Investigator

The project, funded by the MIT-Netherlands Lockheed Martin Seed Fund, developed a set of joint human-AI methods for product design which combined deep generative machine learning methods with the objective functions and design requirements from human designers. It combined the Performance Augmented Diverse Generative Adversarial Networks (PaDGAN) framework developed by Prof. Ahmed’s DeCoDe Lab, with the innovation generation and crowd team methods developed by Dr. Lykourentzou’s Collaborative Technologies Lab. The primary use case of the project was generating innovative bicycle designs, with further applications on high-performance designs for aircraft and airfoil products.

Timeline: 2023 - 2024
Partners: MIT DeCode lab, Utrecht University
Keywords: human-AI collaboration, deep generative models, design optimization, collaborative technologies, machine learning in design, innovation generation 

CROSSCULT
Empowering reuse of digital cultural heritage in context-aware crosscuts of European history

Project Coordinator

The CROSSCULT research project was part of to the "Reflective Societies: Cultural Heritage and European Identities"  topic of European Commission's Horizon 2020 programme. The project consisted of 11 European institutions and 14 associated partners, from the IT, History and Cultural Heritage sectors. Our goal is to help lower cultural barriers and enable the European public to reflect on its common identity, by connecting cultural digital resources, citizen viewpoints and physical venues through technology.

Timeline: 2016 - 2019
Partners:  University College London, French National Center for Scientific Research-CNRS, University of Malta, Centre Virtuel de la Connaissance sur l'Europe, University of Vigo, University of Peloponnese, Technological Educational Institute of Athens, University of Padova, GVAM Guías Interactivas S.L., The National Gallery London
​Keywords: Smart cities, smart venues, reflective societies, citizen science

aCCoRdO
Computational Methods for Human Use Optimization in Complex Crowdsourcing 

Principal Investigator

The aCCoRdO project was a scientific collaboration with the Human-Computer Interaction Institute, Carnegie Mellon University. The project focused on the topic of advanced computational techniques to optimize the involvement of humans in complex crowdsourcing environments. The project was funded by the National Research Fund of Luxembourg (FNR).​

Timeline: 2015 - 2016
Partner: Human-Computer Interaction Institute, Carnegie Mellon University
Keywords: Crowdsourcing, distributed teams

Experimedia BLUE
Personalized Museum Exhibit Monitoring and Description based on Individual Visiting and Cognitive Style

Project Leader

Experimedia BLUE, an FP7 European Commission project, focused on the improvement of QoE (Quality of Experience) of museum visitors, by offering personalized routing and exhibit descriptions tailored to the visitors' visiting styles and cognitive preferences. A range of on-line and on-site technologies were used for this purpose including social network games, indoor localization, scheduling and personalized recommendations. The project involved real users, in a high-technology museum (Foundation of the Hellenic World) in Greece.

Timeline: 2012 - 2013
Partner: University of Peloponnese
Keywords: Future Media Internet, Quality of Experience 

RHEA
Collective Intelligence based Algorithms for the Improvement of Corporate Functions with Application on Operational Risk Management 

Principal Investigator

RHEA, a project funded by the Luxembourg National Research Fund (FNR) under the Marie-Curie COFUND programme, focused on improving knowledge production in corporate wikis by designing algorithm-based mechanisms that combine the collective intelligence of the corporate crowd with machine learning and resource allocation techniques. The application field of the project was operational risk management.

Timeline: 2012 - 2014
​Partner: INRIA - Nancy Grand Est
Keywords: corporate wikis, knowledge harnessing, operational risk management