PhD in Low-Resource Natural Language Processing & Computational Modeling of Language Learning (1.0)

rug university groningen
  • Groningen
  • University of Groningen
Posted on

Applications are invited for a 4-year salaried PhD position within the research project “Polyglot Machines: Human-like Learning of Morphologically Rich Languages”, financed by a NWO-VIDI Talent Grant and coordinated by Principal Investigator (PI) dr. Arianna Bisazza. This is an interdisciplinary project at the intersection of Computational Linguistics/Natural Language Processing (NLP), Computational Psycholinguistics and Language Acquisition.

The PhD Project
Despite the impressive advances made possible by neural networks, current NLP systems are still far from displaying the learning abilities of humans in many languages. By contrast, children around the world acquire extremely diverse languages in comparable time spans and from considerably less linguistic input than that required by neural networks.

This project aims to improve language modeling for low-resource morphologically rich languages, taking inspiration from child language acquisition insights. Among other methodologies, an artificial language learning paradigm will be used to simulate the learning of typologically diverse languages and evaluate the effect of known properties of child-directed language on the acquisition of morphology and other language aspects. Other possible research directions include: the design of better input segmentation methods; language acquisition inspired curriculum learning; and leveraging existing language resources (like dictionaries or morphological analyzers) to boost the learning process in very low-resource settings.

This PhD project offers a unique opportunity to work in an international environment and to acquire valuable research experience: You will be carrying out research in the context of the Computational Linguistics group, which is part of the Centre for Language and Cognition of the University of Groningen (CLCG).

You will be asked to:
- carry out research, present your results and author scientific articles on the above mentioned topics
- collaborate with other members of the team, including postdocs and other PhD candidates
- review Language Acquisition literature relevant to the project’s goals
- engage and collaborate with Language Acquisition researchers interested in the simulation of human language learning with neural networks
- gain teaching experience.

Since its foundation in 1614, the University of Groningen has established an international reputation as a dynamic and innovative university offering high-quality teaching and research. Its 31,000 students are encouraged to develop their own individual talents through challenging study- and career paths. The University of Groningen is an international centre of knowledge: It belongs to the best research universities in Europe and is allied with prestigious partner universities and networks worldwide.

The Faculty of Arts is a large, dynamic faculty in the heart of the city of Groningen. It has more than 5000 students and 700 staff members, who are working at the frontiers of knowledge every day. The Faculty offers a wide range of degree programmes: 15 Bachelor's programmes And over 35 Master's specialisations. Our research, which is internationally widely acclaimed, covers Archaeology, Cultural Studies, History, International Relations, Language and Literary Studies, Linguistics and Media and Journalism Studies.

Ole Gmelin

"Speaking Dutch in the Netherlands will always open certain doors for you when it comes to finding a job. But, it isn’t always a must."

Will you become our new PhD in Low-Resource Natural Language Processing & Computational Modeling of Language Learning (1.0)? Apply at University of Groningen