My name is Ahmet Üstün and I’m PhD Student in the Center for Language and Cognition (CLCG) at the University of Groningen. I’m working as a member of Computational Linguistics research group under the supervision of Arianna Bisazza, Gosse Bouma and Gertjan van Noord.
My research focuses on multilingual natural language processing with a special interest of cross-lingual transfer learning. In this context, I worked on cross-lingual word embeddings, multilingual dependency parsing and multilingual unsupervised NMT. My reserarch aim is to find efficient multilingual adaptation methods for low-resource languages without suffering the curse of multilinguality. My other research interests include NLP methods for morphologically-rich languages and unsupervised learning of morphology.
Ph.D. in Computational Linguistics, University of Groningen
M.S. in Cognitive Science, Middle East Technical University
B.S. in Electrical and Electronics Engineering, Middle East Technical University
Multilingual Unsupervised Neural Machine Translation with Denoising Adapters
Ahmet Üstün, Alexandre Berard, Laurent Besacier, Matthias Gallé
In EMNLP 2021 [pdf]
Massive Choice, Ample Tasks (MACHAMP): A Toolkit for Multi-task Learning in NLP
Rob van der Goot, Ahmet Üstün, Alan Ramponi, Ibrahim Sharaf, Barbara Plank
In EACL 2021 [outstanding paper award, demo track] [pdf] [website]
UDapter: Language Adaptation for Truly Universal Dependency Parsing
Ahmet Üstün, Arianna Bisazza, Gosse Bouma and Gertjan van Noord
In EMNLP 2020 [pdf] [code]
Characters or Morphemes: How to Represent Words?
Ahmet Üstün, Murathan Kurfalı, and Burcu Can
In 3rd Workshop on Representation Learning for NLP, @ACL2018
[best paper award] [pdf] [code, data]
Sep
Research visit at NAVER LABS Europe
Aug
Our paper “Multilingual Unsupervised Neural Machine Translation with Denoising Adapter” is accepted to EMNLP 2021
Jul
Our paper “On the Difficulty of Translating Free-Order Case-Marking Languages” is accepted to TACL
Jan-Jun
Reseach internship at NAVER LABS Europe on using adapters for multilingual unsupervised NMT under supervision of Alexandre Bérard, Laurent Besacier and Matthias Gallé
Jun
Published paper on zero-shot slot and intent detection with auxiliary tasks in NAACL 2021
Apr
MACHAMP (A toolkit for multi-task learning in NLP) won best paper award in demo track at EACL 2021
Apr
Published paper on effectiveness of dataset embeddings in AdaptNLP at EACL 2021
Sep
UDapter that introduces a multilingual parser with language adapters got accepted to EMNLP 2020
Jul
Published journal paper on unsupervised learning of morphology in Natural Language Engineering
Jul
Published paper on code-switched sentiment analysis for SemEval-2020 Task-9
Jun
MACHAMP (A toolkit for multi-task learning in NLP) is on the Arxiv
Jan
Presented at CLIN 30
Dec
Research visit (2 weeks) at Computer Science department of ITU Copenhagen
Nov
Published paper on transfer learning for event detection at CLIC-it 2019
Sep
Published paper on morphology aware cross-lingual word embeddings at RANLP 2019
Sep
Presented at RANLP 2019
Aug
Published paper at SIGMORPHON 2019 shared task (ACL 2019)
Aug
Presented at SIGMORPHON 2019 (ACL 2019)
Jul
Attended LxMLS 2019
Nov
Started PhD in Computational Linguistics at the University of Groningen
Jul
Published paper [best paper] at RepL4NLP 2018 (ACL 2018)
Mar
Presented at TextLink 2018 Final Action Conference
Nov
Research internship (2 Months) at ILCC at the University of Edinburgh
Sep
Published M.S. Thesis on probabilistic learning of Turkish morphosemantics
Sep
Awarded an entrepreneurship grant by the Technological Research Council of Turkey
Apr
Published paper at CICLing 2017 on bayesian modelling of morphology
Apr
Presented at CICLing 2017
Oct
Published paper on unsupervised morphological segmentation at SLSP 2017
Oct
Presented at SLSP 2017
Apr
Published paper on Turkish PoS tagging at CICLing 2016