Nicole Peinelt

Nicole Peinelt


Former Doctoral Student

Cohort year


Partner Institution


Nicole is a PhD candidate at the Department of Computer Science at the University of Warwick, with a special interest in Natural Language Processing (NLP). She is the organiser of the weekly NLP reading group at The Alan Turing Institute. Prior to undertaking her PhD, Nicole studied Computational Linguistics and Chinese language in Germany, China and Taiwan.

Research interests

Nicole's current research focuses on deep learning models for natural language understanding problems, such as semantic similarity, paraphrase detection and question answering. During her PhD, she identified common dataset biases, developed more robust evaluation metrics and designed novel deep learning models. She is particularly interested in applications to real-world problems.

Achievements and awards


  • Nicole Peinelt, Marek Rei, Maria Liakata (2020): GiBERT - Introducing Linguistic Knowledge into BERT through a Lightweight Gated Injection Method. arXiv preprint: 2010.12532. [PDF]
  • Nicole Peinelt, Dong Nguyen, Maria Liakata (2020): tBERT - Topic Models and BERT Joining Forces for Semantic Similarity Detection. Proceedings of ACL 2020, pages 7047-7055. [PDF]
  • Nicole Peinelt, Dong Nguyen, Maria Liakata (2020): Better Early than Late: Fusing Topics with Word Embeddings for Neural Question Paraphrase Identification. arXiv preprint: 2007.11314. [PDF]
  • Nicole Peinelt, Maria Liakata, Dong Nguyen (2019): Aiming beyond the Obvious - Identifying Non-Obvious Cases in Semantic Similarity Datasets. Proceedings of ACL 2019, pages 2792 - 2798. [PDF]
  • Nicole Peinelt, Maria Liakata, Shu-Kai Hsieh (2017): ClassifierGuesser - A Context-based Classifier Prediction System for Chinese Language Learners. Proceedings of IJCNLP 2017 (System Demonstrations), pages 41-44. [PDF]
  • Chieh-Yang Huang, Nicole Peinelt, Lun-Wei Ku (2016): Automatically Suggesting Example Sentences of Near-Synonyms for Language Learners. Proceedings of COLING 2016 (System Demonstrations), pages 302-306. [PDF]