Junior Researcher in Natural Language Processing: LLM and cultural products in Galician
We are looking for a person interested in collaborating in the COMPEL project, focused on the development of Natural Language Processing (NLP) resources applied to challenging use cases in the analysis of Galician cultural production and its comparison to other European languages.
Objective
Recent NLP technologies like Large Language Models (LLM) have brought major progress in automatic text understanding. However, some scenarios still present challenges, like historical language varieties. Some tasks are also still challenging and less commonly studied, such as the automatic analysis of literary texts. The challenge is even higher in the case of low-resource languages like Galician.
In this context, the COMPEL project is developing NLP-based resources for the analysis of 19ᵗʰ century literature in Galician and for comparing it to other European traditions. The goal is promoting smarter access to cultural heritage and cultural products in Galician.
The position proposed will contribute to project tasks such as the following:
• Corpus development and contributing to the development of related NLP tasks (e.g. historical text normalization)
• Experiment design and execution: Training monolingual or multilingual language models for different tasks and comparing models from different paradigms (classical transfer learning based on pre-trained language models vs. LLMs)
• Participating in publications based on the above research
• Web application development (APIs, front-end) for public access to project resources
Knowledge and skills useful for this position
• Degree in computer engineering or related areas, or an area related to Natural Language Processing (NLP) / Computational Linguistics
• Knowledge of NLP tasks relevant for the research described above
• Knowledge of common libraries in NLP (e.g. Hugging Face transformers, torch)
• Knowledge of Python with respect to the above tasks
Offer details
• Duration: 1 year
• The work can be carried out on a full-time or part-time basis, depending on the situation of the interested individuals (it would be possible to do this work alongside studies).
• Remuneration depends on education and working time and follows USC’s salary grid (https://imaisd.usc.es/ferramentas/calculadora/calculadoracontratos.asp)
• Support for attending conferences based on the research developed at the project is available
• The position is based at CiTIUS at Universidade de Santiago de Compostela. This is a young, dynamic and multidisciplinary environment at a recognized research center in AI.
Applications and contact
If interested, please send your CV and a short motivation letter to:
For questions or further information, please contact researcher Pablo Ruiz Fabo at the same email address.