Research Focus
Neural Machine Translation
Developing advanced NMT architectures with a focus on low-resource languages and specialized domains.
Large Language Models
Investigating LLMs' capabilities in translation tasks and exploring efficient fine-tuning methods.
Creative-Text Translation
Research on improving the quality of creative-text translation through data collection, fine-tuning, and development of evaluation metrics.
Education
PhD in Artificial Intelligence
2022 - PresentUniversity of Pisa
Machine Translation & Natural Language Processing
- Research focus on Creative MT and LLMs
- Teaching assistant for Computational Linguistics
- Published papers in top conferences
- Collaborating with industry partners on MT projects
MSc in Linguistics
2020 - 2022University of Naples 'L'Orientale'
Computational Linguistics
- Graduated with honors
- Thesis on NMT and Evaluation Metrics
- Developed an evaluation prototype for Chinese
Selected Publications
Prompting Large Models for Idiomatic Translation
Proceedings of the First Workshop on Creative-Text Translation (2024)
An exploration of LLMs' capabilities in translating idiomatic expressions, focusing on the EN-IT language pair and investigating the impact of prompt design on translation quality. We find that the quality of the translation is highly dependent on the prompt used, as well as the natural language understanding of the LLM.
The SETU-DCU Submissions to IWSLT 2024 Low-Resource Speech-to-Text Translation Task
Proceedings of the International Workshop on Speech Language Translations (2024)
The paper details the SETU-DCU submissions to the IWSLT 2024 Low-Resource Speech-to-Text Translation Task, where we participated with one cascaded system and one end-to-end system on the low-resource pair English-Gaelic. The cascaded system uses Whisper for ASR and mBART/NLLB for MT, while the end-to-end system uses a fine-tuned Conformer model.
Featured Projects
View All
Developer Tools Collection
A suite of web-based tools for developers, featuring color palette generation, CSS gradient creation, and SVG wave design.
View Project