👋 I’m Bram and I work on natural language processing and machine translation (evaluation) and I put a lot of emphasis on open science and open-source contribution.
🤖 I am a post-doctoral researcher in two academic institutions. I work on machine translation evaluation at Ghent University thanks to a grant acquired from CLARIN.eu for the MATEO project (MAchine Translation Evaluation Online). I am also affiliated with KU Leuven where I work on the Horizon 2020 project SignON, which aims to make strides in machine-translating to and from sign language. Within this European project, I am focused on interlingual representations through meaning extraction and language modeling.
🌌 In addition to my academic adventures, I also spend much of my time in the open-source universe. As a Hugging Face core contributor, I spent my time in the trails of repositories such as transformers and evaluate, but I also frequent common NLP libraries such as spaCy, stanza and others. In the spirit of open scienec, I also publish my own code on GitHub, for instance spacy-conll, an integration with spaCy to output CoNLL-U formats and astred, a research framework that calculates syntactic similarity between two sentences and includes word alignment functionality.
🔓 Some of the datasets, models, and demos that I have built are available on the Hugging Face hub, including Dutch datasets for building conversational models, models and a demo to extract meaning from multilingual texts, and models for sentiment analysis in Dutch.