Scholar
David Vos
Google Scholar ID: KvMDZ_QAAAAJ
PhD Student IRLab Amsterdam
Citations & Impact
Citations
54
H-index
3
i10-index
2
Publications
4
Co-authors
4
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'Reliable and Controllable Learned Sparse News Recommendations for Short-History Users' under review.
- Paper 'A Study of Normative Diversity Metrics in News Recommendations' under review.
- Paper 'Parameter-Efficient Automation of Data Wrangling Tasks with Prefix-Tuning' recognized as Best Paper Runner-up at The Table Representation Learning Workshop @ NeurIPS 2022.
- Paper 'Revisiting Language Models in Neural News Recommender Systems' to be presented at ECIR 2025.
Research Experience
- Lectures on text representation learning, sequential recommender systems, and transformers at the University of Amsterdam.
- Participated in the Search Futures Workshop, discussing the role of LLMs in democratic news recommendation systems.
- Presented at the NeurIPS 2022 TRL Workshop on prefix-tuning for data wrangling.
- Gave a talk for the ING Data Science team on LLMs for data management.
- Presented at the Dutch Seminar on Data Systems Design on prefix-tuning for data wrangling.
Education
- Master's degree: Artificial Intelligence, University of Amsterdam (Cum laude)
- Minor: Computer Science, University of Edinburgh
- Bachelor's degree: Artificial Intelligence, University of Amsterdam (Cum laude, Honours programme)
Background
- Research interests include representation learning, generative modeling, and ranking for personalized recommendations and search. Works at the Information Retrieval Lab Amsterdam, collaborating with the Responsible Media Lab to apply research at DPG Media.
Miscellany
- Contact: d.j.a.vos@uva.nl
- Google Scholar, Github, LinkedIn
