Trevor Cohn
Professor

Level 4, Melbourne Connect

Computing and Information Systems
The University of Melbourne
Natural Language Processing group

Ph: +61383441732

Understanding of human language by computers has been a central goal of Artificial Intelligence since its beginnings, with massive potential to improve communication, provide better information access and automate basic human tasks. My research focuses on technologies for automatic processing of human language, with several applications including automatic translation. My core focus is on probabilistic machine learning modelling of language applications, particularly handling uncertain or partly observed data and structured prediction problems.

News

• April 2023-: I now work for Google Research Australia in Sydney. This is a dual appointment, however the majority of my time is at Google, and I will not be taking on new students nor teaching etc.
• October 2022: Thinh Truong awarded AACL award for best paper runner-up
• July 2022: Nitika Mathur awarded CORE dissertation award
• February 2022: Appointed Director of the ITTC on medtech, taking over from Tim Baldwin
• July 2020: Nitika Mathur was awarded a best paper at ACL 2020 (honorable mention), for Tangled Up in BLEU - Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics
• October 2019: I am travelling to Hong Kong for EMNLP, to present two papers in the main conference and two workshop papers.
• September 2019: I will serve as PC chair for EMNLP 2020, alongside Yang Liu and Yulan He.
• July 2019: The group has papers at ICML, WWW, NAACL, ICASSP and ACL (x3).
• July 2018: We hosted ACL 2018 in Melbourne, with Tim Baldwin, Karin Verspoor and myself serving as the local chairs.

Recent Papers

• Xuanli He, Qiongkai Xu, Jun Wang, Benjamin Rubinstein and Trevor Cohn (2024). SEEP: Training Dynamics Grounds Latent Representation Search for Mitigating Backdoor Poisoning Attacks. In Trans. Assoc. Comput. Linguistics, Vol 12.
• Zheng Lim, Harry Stuart, Simon Deyne, Terry Regier, Ekaterina Vylomova, Trevor Cohn and Charles Kemp (2024). A Computational Approach to Identifying Cultural Keywords Across Languages. In Cogn. Sci., Vol 48.
• Jun Wang, Qiongkai Xu, Xuanli He, Benjamin Rubinstein and Trevor Cohn (2024). Backdoor Attacks on Multilingual Machine Translation. In NAACL 2024.
• Thinh Truong, Yulia Otmakhova, Karin Verspoor, Trevor Cohn and Timothy Baldwin (2024). Revisiting subword tokenization: A case study on affixal negation in large language models. In NAACL 2024.
• Fan Jiang, Tom Drummond and Trevor Cohn (2024). Pre-training Cross-lingual Open Domain Question Answering with Large-scale Synthetic Supervision. In EMNLP 2024.
• Zheng Lim, Ekaterina Vylomova, Trevor Cohn and Charles Kemp (2024). Simpson's Paradox and the Accuracy-Fluency Tradeoff in Translation. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics.

• Thinh Truong, Timothy Baldwin, Karin Verspoor and Trevor Cohn (2023). Language models are not naysayers: an analysis of language models on negation benchmarks. In *SEM@ACL 2023.
• Chunhua Liu, Trevor Cohn and Lea Frermann (2023). Seeking Clozure: Robust Hypernym extraction from BERT with Anchored Prompts. In *SEM@ACL 2023.
• Zhuohan Xie, Trevor Cohn and Jey Lau (2023). The Next Chapter: A Study of Large Language Models in Storytelling. In INLG 2023.
• Biaoyan Fang, Trevor Cohn, Timothy Baldwin and Lea Frermann (2023). It's not only What You Say, It's also Who It's Said to: Counterfactual Analysis of Interactive Behavior in the Courtroom. In Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics.
• Xudong Han, Timothy Baldwin and Trevor Cohn (2023). Everybody Needs Good Neighbours: An Unsupervised Locality-based Method for Bias Mitigation. In ICLR 2023.
• Zhuohan Xie, Miao Li, Trevor Cohn and Jey Lau (2023). DeltaScore: Fine-Grained Story Evaluation with Perturbations. In EMNLP.
• Fan Jiang, Qiongkai Xu, Tom Drummond and Trevor Cohn (2023). Boot and Switch: Alternating Distillation for Zero-Shot Dense Retrieval. In EMNLP.
• Fan Jiang, Tom Drummond and Trevor Cohn (2023). Noisy Self-Training with Synthetic Queries for Dense Retrieval. In EMNLP.
• Xuanli He, Qiongkai Xu, Jun Wang, Benjamin Rubinstein and Trevor Cohn (2023). Mitigating Backdoor Poisoning Attacks through the Lens of Spurious Correlation. In EMNLP 2023.
• Biaoyan Fang, Trevor Cohn, Timothy Baldwin and Lea Frermann (2023). More than Votes? Voting and Language based Partisanship in the US Supreme Court. In EMNLP.
• Qiongkai Xu, Trevor Cohn and Olga Ohrimenko (2023). Fingerprint Attack: Client De-Anonymization in Federated Learning. In ECAI 2023.
• Viktoria Schram, Daniel Beck and Trevor Cohn (2023). Performance Prediction via Bayesian Matrix Factorisation for Multilingual Natural Language Processing Tasks. In EACL 2023.
• Shima Khanehzar, Trevor Cohn, Gosia Mikolajczak and Lea Frermann (2023). Probing Power by Prompting: Harnessing Pre-trained Language Models for Power Connotation Framing. In EACL 2023.
• Fan Jiang, Tom Drummond and Trevor Cohn (2023). Don't Mess with Mister-in-Between: Improved Negative Search for Knowledge Graph Completion. In EACL 2023.
• Xudong Han, Timothy Baldwin and Trevor Cohn (2023). Fair Enough: Standardizing Evaluation and Model Selection for Fairness Research in NLP. In EACL 2023.
• Terry Zhuo, Qiongkai Xu, Xuanli He and Trevor Cohn (2023). Rethinking Round-Trip Translation for Machine Translation Evaluation. In ACL.
• Qin Zhang, Shangsi Chen, Dongkuan Xu, Qingqing Cao, Xiaojun Chen, Trevor Cohn and Meng Fang (2023). A Survey for Efficient Open Domain Question Answering. In ACL 2023.
• Zheng Lim, Trevor Cohn, Charles Kemp and Ekaterina Vylomova (2023). Predicting Human Translation Difficulty Using Automatic Word Alignment. In ACL.
• Sayantan Dasgupta, Trevor Cohn and Timothy Baldwin (2023). Cost-effective Distillation of Large Language Models. In ACL.