Research
(to be updated with more info)
My research currently focuses on the area of Social Machine Learning: How to apply Data Mining and Machine Learning techniques in order to assist solving social issues in different areas of our life.
I am currently working on three themes and diverse projects:
1)Machine learning for the people:
Machine learning has brought a revolution to our lives and has more and more applications in different domains. The goal is to build responsible and interpretable models that solve current societal issues and automate/improve current processes.
Current projects:
- VOXReality (HORIZON-CL4-2021-HUMAN-01 Research and Innovation Project) Voice driven interactions in eXtended Reality spaces
- NWO Gravitation Grant: New Science of Mental Disorders
- Beating the Binge
- Digital Legal Studies: Building NLP systems to automate legal processes
Previous projects:
- EIT/EIT Health Initiative: Combating Childhood Obesity
- #MeTooMaastricht: Building a chatbot to assist survivors of sexual harassment
- INTERREG/IMPACT: Artificial Intelligence and Arts: Infinity Games
- NWO/Phillips: “Train your brain, ThinkSlim”
2)Computational Social Media:
Internet and Social Media have not only changed the way many people communicate but also have disrupted the way that consumers, businesses and regulators should think and work. In this context, I am interested into harnessing the power of social media data to explore modern online harms (e.g. misleading advertising, content moderation, complaints, reviews) and understand how people react to and how they use these new technologies.
Current projects:
- Advertising and influencer marketing (HumanAds)
- Data Markets on the Dark Web
- Dark Patterns and Unsafe products/Complaints on the Internet
3)Structuring the Unstructured:
Collecting (e.g. by scrapping) and aggregating (e.g. from different sources) information (mostly unstructured: text, images, measurements etc.) from the Web (or other systems) is a promising direction with many applications:
- Text understanding: Efficiently organizing large document collections (e.g. in the legal domain or short text from social media) by taking into account their underlying semantic information and tracking their evolution over time.
- Domain-specific applications: Lots of domains are in need of exploring the underlying structures of data that are generated. That includes the higher education area (learning analytics, exploring embedding techniques for automating assessment etc.) and the air traffic management system (analyzing the structure of airline/airport networks, predicting links and bottlenecks in the complex aviation system etc.)