Last Friday, I spent 24 hours with my colleagues Hugo Huurdeman, Richard Zijdeman, and Marieke van Erp in a church in Amsterdam to work on an idea that I had been toying around with some time. The idea being a repository of historical recipes. However, getting your hands on historical recipes is notoriously hard. The newspapers in Delpher, the Dutch newspaper corpus, contain many recipes. We extracted and enriched these recipes using NLP, Machine Learning, and LOD during the HackaLOD. We are currently improving the prototype which we will hopefully present at DH Benelux 2018.
More on the HackaLOD here