I’m a research scientist at Google Research, Montreal, where I work on improving machine translation. I’m generally interested in the intersection of natural language processing (NLP) and machine learning.
In my PhD (Heidelberg University, Germany) I investigated how reinforcement learning algorithms can be used to turn weak supervision signals from users into meaningful updates for a machine translation system.
One of my long-term goals is to make research in NLP more accessible, along multiple dimensions:
- Underresourced NLP: Produce research outcomes for underresourced languages, such that not only English-speaking users can benefit from the progress we’re making in NLP.
- Novices: Reduce the entry burdens (in terms of coding and research practices) for novices in the field, especially for new students or researchers from other related areas.
- Science outreach: Get the general public, especially highschool students, more interested in research in machine learning to grow a better understanding of what our current methods look like and where their limitations are.
- Oct 2020:
- Paper accepted at COLING 2020. See here.
- Sep 2020:
- Paper accepted at EMNLP 2020. Summarizing my internship project at Google, this paper analyses generation orders for mask-based non-autoregressive translation. Preprint coming.
- Paper accepted at EMNLP Findings 2020. Follow-up on the AfricaNLP paper about the Masakhane project. Preprint coming.
- My thesis is officially published, you can download it here.
- Apr 2020:
- Paper accepted at EAMT 2020. This paper concludes my PhD thesis and empirically measures the trade-off between (human) feedback strength and model improvement in machine translation. Preprint here.
- Mar 2020:
- Successfully defended my PhD thesis.
- Papers accepted at the AfricaNLP workshop at ICLR.
- The first paper describes the efforts of the Masakhane community to build machine translation models for as many African languages as possible, by growing and fostering a distributed community of African researchers, students, and computer scientists (and more). Preprint here. Check out the Masakhane GitHub repo for the benchmarks for details.
- The second paper describes the tuning of Transformer model depth for low-resource machine translation. Preprint here.
- Feb 2020: I joined the Google Translate team in Montreal <3.
- Jan 2020: Handed in my thesis, finally.