Postdoctoral researcher position in machine learrning and natural language processing (1 year, Université Paris-Saclay, France)

[Caio Filippo Corro]

Hello,

we have a position available for a 1 year postdoc at Université Paris-Saclay to work on neural architectures for natural language processing, see below.

If interested, contact us be e-mail:
- Caio Corro caio....@limsi.fr
- Nadi Tomeh to...@lipn.fr

Best,
Caio

Postdoctoral researcher in Machine Learning and Natural Language Processing

Mission
The postdoc will be part of the TLP team of the \textit{Laboratoire Interdisciplinaire des Sciences du Numérique} (LISN, formerly LIMSI) located at the University of Paris-Saclay and will be supervised by Caio Corro (TLP, LISN) and Nadi Tomeh (RCLN, Laboratoire Informatique de Paris Nord). The recruted postdoc will be in charge of designing neural architectures for natural language processing. In particular, we will be interested in modeling robustness to certain linguistic phenomena such as reordering.
Activities

- Study previous work on neural architectures motivated by expert knowledge, see for example "Group Equivariant Convolutional Networks" (Cohen, Welling)
- Propose new neural architectures motivated by language syntax in monolingual and multilingual settings
- Evaluate the proposed neural architectures on syntactic and/or semantic analysis tasks
- Participate in the supervision of two Master internships on the project's theme

The postdoc will be encouraged to publish in top-ranked conferences (NeurIPS/ICML/ICLR and *ACL/EMNLP).

Skills

The candidate should have a PhD in machine learning and/or natural language processing. He/she should have a very good understanding of modern neural architectures commonly used in NLP (LSTM, attentional networks/transformers) and should show an interest in syntax and language typology.

The candidate is expected to have good writing and communication skills in English.

Work context

The work will be carried out in the TLP team of LISN, located in the south of Paris, in collaboration with the RCLN team of LIPN (Northern Paris Computer Science Lab).

The LISN hosts two teams specialized in NLP including different modalities (written languages, spoken languages and sign languages) in a multidisciplinary approach (i.e. researchers in computer science and linguistics). In particular, part of the TLP team in which the candidate will work is specialized in machine learning, machine translation and syntactic analysis.

The candidate will also be an associate member of LIPN and will join the RCLN team, a member of the laboratory of excellence network EFL (Empirical Foundations of Linguistics). The research themes of RCLN are organized around syntactic and semantic analysis, but also annotation and knowledge acquisition from texts.


The researcher will benefit from an environment adapted to the pursuit of his research (equipment and access to several computing clusters). In particular, the team has access to a very large number of NVidia V100 GPUs via the Jean-Zay supercomputer.