Research interests and activities

My main research interests are diverse. At the highest levels I am interested in information retrieval and data mining theoretics and applications. I started with some projects and a master’s thesis in Hidden Markov Models and imprecise probability theory in speech and text processing1, and I also did a project on music information retrieval. But for my PhD2 I went to social media text mining. Very quickly I began to look at deep learning techniques since, according to literature, this seemed the way to further advance the field of natural language processing. Next to that I am also interested in applications in recommendation systems and education. Currently I am working on the topic of text similarity and language models using word embeddings. I have counseled master’s dissertations on the topic of popularity prediction for tweets and news articles, style transfer in music composition, POI recommendation using social media, and large-scale sentiment classification of tweets.

Academic papers

Below is a list of published and unpublished academic work. More information can also be found on my Google Scholar profile.


  • Best paper award at the fifth Making Sense of Microposts workshop at WWW 2015, Florence, Italy.

  1. Under supervision of professor Gert de Cooman, PhD and counseled by Jasper de Bock, PhD and Arthur Van Camp.

  2. Under supervision of professor Bart Dhoedt, PhD and Thomas Demeester, PhD. Formerly funded by BOF Ghent University, currently funded by FWO Flanders.