Hi, my name is Fréderic Godin. I am a Ph.D. researcher at the IDLab of Ghent University, located in the beautiful city of Ghent, Belgium. My research interests are centered around noisy social media data and machine learning. I am particularly interested in natural language processing/understanding (NLP/NLU) of microposts using representation learning and (deep) neural networks. Recently, I also did my first Computer Vision project via a Kaggle competition.

Research Interests

Currently my main research interest is the use of representation learning and/or deep learning techniques for natural language processing (NLP) of microposts such as tweets. These microposts are very noisy and therefore current techniques that rely on manual feature engineering are less suited.

I’ve also participated in the $1,000,000 Lung Cancer Prediction challenge on Kaggle where we applied several types of neural network architectures on 3D CT scans of lungs.

Apart from that, I have built several NLP applications such as event detection in soccer games, predicting the outcome of soccer games (together with Jasper Zuallaert), and a hashtag recommendation system using topic models. Together with some Flemish broadcasters, I also looked into the correlation between TV viewing rates and Twitter traffic.

Academic Track

2012 – now : Ph.D. @ Ghent University funded with an IWT Fellowship
2010 – 2012 : Master of Science in Engineering, Computer Science, Summa Cum Laude @ Ghent University
2007 – 2010 : Bachelor of Science in Engineering, Computer Science, Cum Laude @ Ghent University

Some more information can be found at my LinkedIn profile. You can contact me via firstname.lastname@ugent.be (no accents on my first name).