Hi, I am Madelon Hulsebos. I have been a data scientist since 2019. Currently, I work for Heineken.

Throughout the last few years, I heard many stories about organizations that struggle to unlock the value in their data due to unaligned data source systems. I also observed it myself. In 2018, at the MIT Media Lab, I got the chance to contribute the method (Sherlock) to tackle semantic type detection for tabular data at scale, which is meant to partly solve this issue.

Stimulated by the rapid advances in the field of NLP, I believe accurately recovering semantics from tabular data is within close reach. Actually, I contribute to this myself as a PhD candidate at the INDE Lab at the University of Amsterdam supervised by Paul Groth and co-supervised by Çağatay Demiralp.

For more background, you can find my CV here, check out the Projects and Posts pages, or my other pages (see below).