Abstract
With the increasing capabilities of large language models (LLMs), these high-performance models have achieved state-of-the-art results on a wide range of natural language processing (NLP) tasks. However, the models’ performance on commonly-used benchmark datasets often fails to accurately reflect their reliability and robustness when applied to real-world noisy data. To address these challenges, we propose a unified robustness evaluation framework based on the slot-filling task to systematically evaluate the dialogue understanding capability of LLMs in diverse input perturbation scenarios. Specifically, we construct a input perturbation evaluation dataset, Noise-LLM, which contains five types of single perturbation and four types of mixed perturbation data. Furthermore, we utilize a multi-level data augmentation method (character, word, and sentence levels) to construct a candidate data pool, and carefully design two ways of automatic task demonstration construction strategies (instance-level and entity-level) with various prompt templates. Our aim is to assess how well various robustness methods of LLMs perform in real-world noisy scenarios. The experiments have demonstrated that the current open-source LLMs generally achieve limited perturbation robustness performance. Based on these experimental observations, we make some forward-looking suggestions to fuel the research in this direction (The code is available at https://github.com/ZhaoJin-xu/A-Unified-Robustness-Evaluation-Framework-for-Noisy-Slot-Filling-Task).
G. Dong and J. Zhao—The first two authors contribute equally.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Brown, T.B., et al.: Language models are few-shot learners. ArXiv abs/2005.14165 (2020)
Brown, T.B., et al.: Language models are few-shot learners (2020)
Coucke, A., et al.: Snips voice platform: an embedded spoken language understanding system for private-by-design voice interfaces (2018)
Dong, G., et al.: PSSAT: a perturbed semantic structure awareness transferring method for perturbation-robust slot filling. In: Proceedings of the 29th International Conference on Computational Linguistics, pp. 5327–5334. International Committee on Computational Linguistics, Gyeongju (2022). https://aclanthology.org/2022.coling-1.473
Gopalakrishnan, K., Hedayatnia, B., Wang, L., Liu, Y., Hakkani-Tur, D.: Are neural open-domain dialog systems robust to speech recognition errors in the dialog history? An empirical study (2020)
Gui, T., et al.: TextFlint: unified multilingual robustness evaluation toolkit for natural language processing (2021)
Guo, D., et al.: Revisit out-of-vocabulary problem for slot filling: a unified contrastive frameword with multi-level data augmentations. ArXiv abs/2302.13584 (2023)
Hu, E.J., et al.: LoRA: low-rank adaptation of large language models (2021)
Lazaridou, A., Gribovskaya, E., Stokowiec, W., Grigorev, N.: Internet-augmented language models through few-shot prompting for open-domain question answering. ArXiv abs/2203.05115 (2022)
Li, X., et al.: A robust contrastive alignment method for multi-domain text classification. In: ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 7827–7831 (2022). https://doi.org/10.1109/ICASSP43922.2022.9747192
Li, X., et al.: Generative zero-shot prompt learning for cross-domain slot filling with inverse prompting. In: Findings of the Association for Computational Linguistics: ACL 2023, pp. 825–834. Association for Computational Linguistics, Toronto (2023). https://aclanthology.org/2023.findings-acl.52
Lin, J., et al.: M6: multi-modality-to-multi-modality multitask mega-transformer for unified pretraining. In: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, KDD 2021, pp. 3251–3261. Association for Computing Machinery, New York, NY (2021). https://doi.org/10.1145/3447548.3467206
Ma, E.: NLP augmentation (2019). https://github.com/makcedward/nlpaug
Moradi, M., Samwald, M.: Evaluating the robustness of neural language models to input perturbations. In: Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pp. 1558–1570. Association for Computational Linguistics, Online and Punta Cana (2021). https://doi.org/10.18653/v1/2021.emnlp-main.117, https://aclanthology.org/2021.emnlp-main.117
Namysl, M., Behnke, S., Köhler, J.: NAT: noise-aware training for robust neural sequence labeling. In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pp. 1501–1517. Association for Computational Linguistics (2020). https://doi.org/10.18653/v1/2020.acl-main.138, https://aclanthology.org/2020.acl-main.138
OpenAI: GPT-4 technical report (2023)
Ouyang, L., et al.: Training language models to follow instructions with human feedback (2022)
Peng, B., Li, C., Zhang, Z., Zhu, C., Li, J., Gao, J.: RADDLE: an evaluation benchmark and analysis platform for robust task-oriented dialog systems. arXiv preprint arXiv:2012.14666 (2020)
Qixiang, G., et al.: Exploiting domain-slot related keywords description for few-shot cross-domain dialogue state tracking. In: Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pp. 2460–2465. Association for Computational Linguistics, Abu Dhabi (2022). https://aclanthology.org/2022.emnlp-main.157
Reimers, N., Gurevych, I.: Sentence-BERT: sentence embeddings using Siamese BERT-networks. arXiv preprint arXiv:1908.10084 (2019)
Ruan, W., Nechaev, Y., Chen, L., Su, C., Kiss, I.: Towards an ASR error robust spoken language understanding system. In: Interspeech 2020 (2020). https://www.amazon.science/publications/towards-an-asr-error-robust-spoken-language-understanding-system
Shen, Y., Song, K., Tan, X., Li, D., Lu, W., Zhuang, Y.: HuggingGPT: solving AI tasks with ChatGPT and its friends in HuggingFace (2023)
Touvron, H., et al.: LLaMA: open and efficient foundation language models (2023)
Wei, J., et al.: Chain-of-thought prompting elicits reasoning in large language models (2023)
Wu, D., Chen, Y., Ding, L., Tao, D.: Bridging the gap between clean data training and real-world inference for spoken language understanding. arXiv preprint arXiv:2104.06393 (2021)
Zeng, W., et al.: Semi-supervised knowledge-grounded pre-training for task-oriented dialog systems. In: Proceedings of the Towards Semi-Supervised and Reinforced Task-Oriented Dialog Systems (SereTOD), pp. 39–47. Association for Computational Linguistics, Abu Dhabi (2022). https://aclanthology.org/2022.seretod-1.6
Zhang, Y., et al.: Pay attention to implicit attribute values: a multi-modal generative framework for AVE task. In: Findings of the Association for Computational Linguistics: ACL 2023, pp. 13139–13151. Association for Computational Linguistics, Toronto (2023). https://aclanthology.org/2023.findings-acl.831
Zhao, G., Dong, G., Shi, Y., Yan, H., Xu, W., Li, S.: Entity-level interaction via heterogeneous graph for multimodal named entity recognition. In: Findings of the Association for Computational Linguistics: EMNLP 2022, pp. 6345–6350. Association for Computational Linguistics, Abu Dhabi (2022). https://aclanthology.org/2022.findings-emnlp.473
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Dong, G. et al. (2023). Revisit Input Perturbation Problems for LLMs: A Unified Robustness Evaluation Framework for Noisy Slot Filling Task. In: Liu, F., Duan, N., Xu, Q., Hong, Y. (eds) Natural Language Processing and Chinese Computing. NLPCC 2023. Lecture Notes in Computer Science(), vol 14302. Springer, Cham. https://doi.org/10.1007/978-3-031-44693-1_53
Download citation
DOI: https://doi.org/10.1007/978-3-031-44693-1_53
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-44692-4
Online ISBN: 978-3-031-44693-1
eBook Packages: Computer ScienceComputer Science (R0)