Koen Verstrepen

Post-doctoral researcher at the Research Group
Advanced Database Research and Modelling (ADReM)
University of Antwerp
Department of Mathematics and Computer Science
Office G.329
Middelheimlaan 1
2020 Antwerpen
Belgium

Tel: (+32) 3 265 32 64
Fax: (+32) 3 265 37 77
E-mail: koen.verstrepen(a)uantwerpen.be



Research Topic


My research is centred around recommender systems. More specifically, I currently focus on collaborative filtering with binary, positive-only data. Applications that correspond to this topic are:

  • recommending videos based on the videos you played in the past,
  • recommending products based on the products you bought in the past
  • recommending websites based on the websites you visited in the past
  • recommending tags for pictures based on the tags already assigned to the picture
  • ...

Publications


Collaborative Filtering with Binary, Positive-only Data
Koen Verstrepen
PhD Thesis, University of Antwerp, 2015.

Top-N Recommendation for Shared Accounts
Koen Verstrepen and Bart Goethals
In Proceedings of the 9th ACM Conference on Recommender Systems (RecSys'15), 2015.

Unifying Nearest Neighbors Collaborative Filtering
Koen Verstrepen and Bart Goethals
In Proceedings of the 8th ACM Conference on Recommender Systems (RecSys'14), 2014.

Implementations


KUNN: KUNN Unified Nearest Neighbors (by Koen Verstrepen and Bart Goethals)
DAMIB-COVER: DAMIB-COVER (by Koen Verstrepen and Bart Goethals)

Tutorials


Collaborative Filtering with Binary, Positive-Only Data

Teaching


Introduction to Recommender Systems (slides).