Dr Nafise Sadat Moosavi
School of Computer Science
Lecturer in Natural Language Processing
Member of the Natural Language Processing research group
Interdisciplinary Research Lead
N.S.Moosavi@sheffield.ac.uk
Regent Court (CS)
Full contact details
Dr Nafise Sadat Moosavi
School of Computer Science
Regent Court (CS)
211 Portobello
Sheffield
S1 4DP
School of Computer Science
Regent Court (CS)
211 Portobello
Sheffield
S1 4DP
- Profile
-
Dr. Nafise Sadat Moosavi is a Lecturer in Natural Language Processing at the Computer Science Department of the University of Sheffield. Before joining the University of Sheffield, she was a postdoctoral researcher at the UKP Lab at the Technical University of Darmstadt. Before that, she got her PhD degree from Heidelberg University. She received her bachelor's and master's degrees in computer science from Alzahra University and the Sharif University of Technology in Tehran, Iran.
- Research interests
-
- Robust and trustworthy large language models
- Reasoning and inference in NLP systems
- Bias, fairness, and responsible AI
- Reliable evaluation of language models
- Culturally adaptive AI
- Efficient and scalable NLP
- Publications
-
Show: Featured publications All publications
Featured publications
This person does not have any publications available.
All publications
Journal articles
- Exploring Gender Disparities in Automatic Speech Recognition Technology.. CoRR, abs/2502.18434.
- Scoring coreference chains with split-antecedent anaphors. Dialogue & Discourse, 14(2), 1-48. View this article in WRRO
- NL-Augmenter . Northern European Journal of Language Technology, 9(1).
Conference proceedings
- How to leverage digit embeddings to represent numbers?. Proceedings of the 31st International Conference on Computational Linguistics (pp 7685-7697). Abu Dhabi, United Arab Emirates, 19 January 2025 - 19 January 2025. View this article in WRRO
- Rolling the DICE on Idiomaticity: How LLMs Fail to Grasp Context. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp 7314-7332), July 2025 - July 2025.
- MultiHoax: A Dataset of Multi-hop False-premise questions. Findings of the Association for Computational Linguistics: ACL 2025 (pp 10169-10187), July 2025 - July 2025.
- Fairness in Automatic Speech Recognition Isn’t a One-Size-Fits-All. Findings of the Association for Computational Linguistics: EMNLP 2025 (pp 19169-19178), November 2025 - November 2025.
- From Input Perception to Predictive Insight: Modeling Model Blind Spots Before They Become Errors. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (pp 34316-34329), November 2025 - November 2025.
- Beyond Hate Speech: NLP’s Challenges and Opportunities in Uncovering Dehumanizing Language. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (pp 26953-26968), November 2025 - November 2025.
- Deconstructing Attention: Investigating Design Principles for Effective Language Modeling.. IJCNLP-AACL (long papers) (pp 708-727)
- Blending Social Interaction Realms: Harmonizing Online and Offline Interactions through Augmented Reality. Proceedings of the 17th International Symposium on Visual Information Communication and Interaction (pp 1-8)
- Spanning the spectrum of hatred detection: a Persian multi-label hate speech dataset with annotator rationales. Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38(16) (pp 17889-17897). Vancouver, Canada, 20 February 2024 - 20 February 2024. View this article in WRRO
- Universal Anaphora: The First Three Years. 2024 Joint International Conference on Computational Linguistics Language Resources and Evaluation Lrec Coling 2024 Main Conference Proceedings (pp 17087-17100)
- LLMs as narcissistic evaluators: when ego inflates evaluation scores. Findings of the Association for Computational Linguistics: ACL 2024 (pp 12688-12701). Bangkok, Thailand, 11 August 2024 - 11 August 2024. View this article in WRRO
- Lessons learned from a Citizen Science project for Natural Language Processing. Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (pp 3594-3608). Dubrovnik, Croatia, 2 May 2023 - 2 May 2023. View this article in WRRO
- Transformers with learnable activation functions. Findings of the Association for Computational Linguistics: EACL 2023 (pp 2382-2398). Dubrovnik, Croatia, 2 May 2023 - 2 May 2023. View this article in WRRO
- FERMAT: an alternative to accuracy for numerical reasoning. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Vol. 1 (pp 15026-15043). Toronto, Canada, 9 July 2023 - 9 July 2023. View this article in WRRO
- Arithmetic-based pretraining improving numeracy of pretrained language models. Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023), Vol. *SEM 2023 (pp 477-493). Toronto, Canada, 13 July 2023 - 13 July 2023. View this article in WRRO
- Lessons Learned from a Citizen Science Project for Natural Language Processing. Eacl 2023 17th Conference of the European Chapter of the Association for Computational Linguistics Proceedings of the Conference (pp 3576-3590)
- Transformers with Learnable Activation Functions. Eacl 2023 17th Conference of the European Chapter of the Association for Computational Linguistics Findings of Eacl 2023 (pp 2337-2353)
- Learning From Free-Text Human Feedback – Collect New Datasets Or Extend Existing Ones?. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (pp 16259-16279), December 2023 - December 2023.
- Layer or representation space: what makes BERT-based evaluation metrics robust?. Proceedings of the 29th International Conference on Computational Linguistics, Vol. 29(1) (pp 3401-3411). Gyeongju, Republic of Korea, 12 October 2022 - 12 October 2022. View this article in WRRO
- Falsesum : generating document-level NLI examples for recognizing factual inconsistency in summarization. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (pp 2763-2776). Seattle, WA, USA, 10 July 2022 - 10 July 2022. View this article in WRRO
- Adaptable adapters. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (pp 3742-3753). Seattle, WA, USA, 10 July 2022 - 10 July 2022. View this article in WRRO
- The Universal Anaphora Scorer. Proceedings of the Thirteenth Language Resources and Evaluation Conference, LREC 2022 (pp 4873-4883). Marseille, France, 20 June 2022 - 20 June 2022. View this article in WRRO
- SciGen : a dataset for reasoning-aware text generation from scientific tables. Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks 1 (NeurIPS Datasets and Benchmarks 2021), 6 December 2021 - 6 December 2021. View this article in WRRO
- Avoiding inference heuristics in few-shot prompt-based finetuning. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing (pp 9063-9074). Punta Cana, Dominican Republic (Online), 7 November 2021 - 7 November 2021. View this article in WRRO
- Coreference reasoning in machine reading comprehension. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Vol. 1 (pp 5768-5781). Bangkok, Thailand (Online), 1 August 2021 - 1 August 2021. View this article in WRRO
- Stay together: a system for single and split-antecedent anaphora resolution. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (pp 4174-4184). Online, 6 June 2021 - 6 June 2021. View this article in WRRO
- Free the plural: unrestricted split-antecedent anaphora resolution. Proceedings of the 28th International Conference on Computational Linguistics (pp 6113-6125). Barcelona, Spain (Online), 8 December 2020 - 8 December 2020. View this article in WRRO
- Improving QA generalization by concurrent modeling of multiple biases. Findings of the Association for Computational Linguistics: EMNLP 2020 (pp 839-853). Online, 16 November 2020 - 16 November 2020. View this article in WRRO
- Towards debiasing NLU models from unknown biases. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020) (pp 7597-7610). Online, 16 November 2020 - 16 November 2020. View this article in WRRO
- Mind the trade-off: debiasing NLU models without degrading the in-distribution performance. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL 2020) (pp 8717-8729). Online, 5 July 2020 - 5 July 2020. View this article in WRRO
- Neural duplicate question detection without labeled training data. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) (pp 1607-1617). Hong Kong, China, 3 November 2019 - 3 November 2019. View this article in WRRO
- COALA : a neural coverage-based approach for long answer selection with small data. Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 33 (pp 6932-6939). Honolulu, HI, USA, 27 January 2019 - 27 January 2019. View this article in WRRO
- Using automatically extracted minimum spans to disentangle coreference evaluation from boundary detection. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL 2019) (pp 4168-4178). Florence, Italy, 28 July 2019 - 28 July 2019. View this article in WRRO
- Using linguistic features to improve the generalization capability of neural coreference resolvers. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing (pp 193-203). Brussels, Belgium, 31 October 2018 - 31 October 2018. View this article in WRRO
- Anaphora resolution with the ARRAU corpus. Proceedings of the First Workshop on Computational Models of Reference, Anaphora and Coreference (pp 11-22). New Orleans, Louisiana, 6 June 2018 - 6 June 2018. View this article in WRRO
- Revisiting selectional preferences for coreference resolution. Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (pp 1332-1339). Copenhagen, Denmark, 9 September 2017 - 9 September 2017. View this article in WRRO
- Lexical features in coreference resolution: to be used with caution. Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, Vol. 2 (pp 14-19). Vancouver, Canada, 30 July 2017 - 30 July 2017. View this article in WRRO
- Use generalized representations, but do not forget surface features. Proceedings of the 2nd Workshop on Coreference Resolution Beyond OntoNotes (CORBON 2017) (pp 1-7). Valencia, Spain, 4 April 2017 - 4 April 2017. View this article in WRRO
- Which coreference evaluation metric do you trust? A proposal for a link-based entity aware metric. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, Vol. 1 (pp 632-642). Berlin, Germany, 7 August 2016 - 7 August 2016. View this article in WRRO
- Search space pruning: a simple solution for better coreference resolvers. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (pp 1005-1011). San Diego, California, 12 June 2016 - 12 June 2016. View this article in WRRO
- Unsupervised coreference resolution using a graph labeling approach. Human Language Technology Challenges for Computer Science and Linguistics, Vol. LNCS 8387 (pp 93-103). Poznań, Poland, 25 November 2011 - 25 November 2011. View this article in WRRO
- Unsupervised coreference resolution by utilizing the most informative relations. Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Papers (pp 644-655). Dublin, Ireland, 23 August 2014 - 23 August 2014. View this article in WRRO
- TCvisor: A hypervisor level secure storage. 2010 International Conference for Internet Technology and Secured Transactions Icitst 2010
Preprints
- No Shortcuts to Culture: Indonesian Multi-hop Question Answering for Complex Cultural Understanding, arXiv.
- RIGOURATE: Quantifying Scientific Exaggeration with Evidence-Aligned Claim Evaluation, arXiv.
- Faithful Summarisation under Disagreement via Belief-Level Aggregation, arXiv.
- LLMs Do Not See Age: Assessing Demographic Bias in Automated Systematic Review Synthesis, arXiv.
- Deconstructing Attention: Investigating Design Principles for Effective Language Modeling, arXiv.
- From Input Perception to Predictive Insight: Modeling Model Blind Spots Before They Become Errors, arXiv.
- Rolling the DICE on Idiomaticity: How LLMs Fail to Grasp Context, arXiv.
- Beyond Hate Speech: NLP's Challenges and Opportunities in Uncovering Dehumanizing Language, arXiv.
- More or Less Wrong: A Benchmark for Directional Bias in LLM Comparative Reasoning, arXiv.
- MultiHoax: A Dataset of Multi-hop False-Premise Questions, arXiv.
- Frame In, Frame Out: Do LLMs Generate More Biased News Headlines than Humans?, arXiv.
- ContrastScore: Towards Higher Quality, Less Biased, More Efficient Evaluation Metrics with Contrastive Evaluation, arXiv.
- Exploring Gender Disparities in Automatic Speech Recognition Technology, arXiv.
- Transforming Science with Large Language Models: A Survey on AI-assisted Scientific Discovery, Experimentation, Content Generation, and Evaluation, arXiv.
- Exploring the Influence of Label Aggregation on Minority Voices: Implications for Dataset Bias and Model Training, arXiv.
- How to Leverage Digit Embeddings to Represent Numbers?, arXiv.
- Decoding News Narratives: A Critical Analysis of Large Language Models in Framing Detection, arXiv.
- LLMs as Narcissistic Evaluators: When Ego Inflates Evaluation Scores, arXiv.
- Learning From Free-Text Human Feedback -- Collect New Datasets Or Extend Existing Ones?, arXiv.
- FERMAT: An Alternative to Accuracy for Numerical Reasoning, arXiv.
- Lessons Learned from a Citizen Science Project for Natural Language Processing, arXiv.
- Scoring Coreference Chains with Split-Antecedent Anaphors, arXiv.
- Arithmetic-Based Pretraining -- Improving Numeracy of Pretrained Language Models, arXiv.
- NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentation, arXiv.
- Improving Robustness by Augmenting Training Sentences with Predicate-Argument Structures, arXiv.
- Improving Generalization by Incorporating Coverage in Natural Language Inference, arXiv.
- Exploring Gender Disparities in Automatic Speech Recognition Technology.. CoRR, abs/2502.18434.
- Grants
-
- Geometric Representations of Uncertainty for Foundation Models, Royal Society, 03/2025 - 03/2028, £193,560, as PI