Scholar
Ehsan Shareghi
Google Scholar ID: EhnQJFwAAAAJ
Monash University
Citations & Impact
Citations
1,628
H-index
17
i10-index
26
Publications
20
Co-authors
17
list available
Publications
Resume (English only)
Academic Achievements
- Supervised students to publish 6 papers at EMNLP 2025
- Student works accepted at COLM 2025, ICML 2025 (Spotlight), NAACL 2025, etc.
- Supervised Michelle’s PhD thesis on 'Advancing Data-Constrained Spoken Language Translation'
- Matthew and Ramya received Outstanding Paper Award at ALTA 2024
- Serving as Senior Area Chair for ACL 2025
- Multiple papers accepted in 2024 at COLM, ACL/Findings, EMNLP/Findings, NAACL, EACL/Findings
- Multiple papers accepted in 2023 at CogSci, EMNLP/EACL-Findings, Interspeech, CoNLL
- Multiple papers accepted in 2022 at ACL, EMNLP, NAACL-Findings, Interspeech
- Multiple papers accepted in 2021 at ACL, EACL, NAACL, EMNLP, RepL4NLP
- Served as Area Chair for EMNLP 2023, EACL 2023, AACL 2023
- Launched the Koala project
- Released Visual Med-Alpaca: Bridging Modalities in Biomedical Language Models
- Co-organized events including Monash GenAI Masterclass, Monash Interdisciplinary Workshops on Language, and DeepLo workshops
Research Experience
- Leads a team of students working on predictive models for language (text and speech)
- Previously worked on integrating nonparametric models and compressed data structures for large-scale text prediction
- Recent work includes multilingual NLP and Bayesian learning of neural models
- Since 2020, has been researching speech and text models for understanding, translation, and reasoning tasks
- Investigates limitations of foundation models and develops model-agnostic solutions leveraging large language and speech models
Background
- Assistant Professor (Senior Lecturer) in the Department of Data Science and Artificial Intelligence, Monash University
- Affiliated Lecturer at the Language Technology Lab, University of Cambridge
- Deputy Director of the Master of AI program at Monash
- Research interests include: augmenting reasoning with LLMs, red teaming safety in foundation models for text and speech, AI applications in law, self-supervised learning, graph-based approaches for knowledge-intensive tasks, deep generative models, and Bayesian learning with small data
- Currently focuses on speech and text models for understanding, translation, and reasoning, especially in low-resource multilingual NLP and Bayesian learning of neural models
