文本分类中Prompt Learning方法研究综述

doi:10.3778/j.issn.1002-8331.2310-0049

摘要/Abstract

摘要： 文本分类是自然语言处理中的一项基础任务，在情感分析、新闻分类等领域具有重要应用。相较于传统的机器学习和深度学习模型，提示学习可以在数据不足的情况下通过构建提示来进行文本分类。近年来，GPT-3的出现推动了提示学习方法的发展，并且在文本分类领域取得了显著的进展。对以往的文本分类方法进行简要梳理，分析其存在的问题与不足；阐述了提示学习的发展进程，以及构建提示模板的方法，并对用于文本分类的提示学习方法研究及成果进行了介绍和总结。最后，对提示学习在文本分类领域的发展趋势和有待进一步研究的难点进行了总结和展望。

关键词: 提示学习, 文本分类, 情绪分析, 新闻分类

Abstract: Text classification is a basic task in natural language processing, which has important applications in sentiment analysis, news classification and other fields. Compared with traditional machine learning and deep learning models, prompt learning can construct prompts for text classification in the case of insufficient data. In recent years, the emergence of GPT-3 has promoted the development of cue learning methods, and has made significant progress in the field of text classification. Firstly, this paper briefly combs the process of previous text classification methods and analyzes their existing problems and shortcomings. Secondly, it expounds the development process of cue learning and the method of constructing cue templates, and introduces and summarizes the research and results of cue learning methods for text classification. Finally, the development trend and difficulties to be further studied in the field of text classification are summarized and prospected.

Key words: prompt learning, text classification, sentiment analysis, news classification

顾勋勋, 刘建平, 邢嘉璐, 任海玉. 文本分类中Prompt Learning方法研究综述[J]. 计算机工程与应用, 2024, 60(11): 50-61.

GU Xunxun, LIU Jianping, XING Jialu, REN Haiyu. Text Classification：Comprehensive Review of Prompt Learning Methods[J]. Computer Engineering and Applications, 2024, 60(11): 50-61.

参考文献

[1] BAYER M, KAUFHOLD M A, REUTER C. A survey on data augmentation for text classification[J]. ACM Computing Surveys, 2022, 55(7): 1-39.
[2] MINAEE S, KALCHBRENNER N, CAMBRIA E, et al. Deep learning--based text classification: a comprehensive review[J]. ACM Computing Surveys, 2021, 54(3): 1-40.
[3] LI Q, PENG H, LI J, et al. A survey on text classification: from traditional to deep learning[J]. ACM Transactions on Intelligent Systems and Technology , 2022, 13(2): 1-41.
[4] WU H, LIU Y, WANG J. Review of text classification methods on deep learning[J]. Computers, Materials & Continua, 2020, 63(3): 1309-1321.
[5] GASPARETTO A, MARCUZZO M, ZANGARI A, et al. A survey on text classification algorithms: from text to predictions[J]. Information, 2022, 13(2): 83.
[6] WANG L, CHEN R, LI L. Knowledge-guided prompt learning for few-shot text classification[J]. Electronics, 2023, 12(6): 1486.
[7] ZHANG P, CHAI T, XU Y. Adaptive prompt learning-based few-shot sentiment analysis[J]. Neural Processing Letters, 2023, 55(6): 7259-7272.
[8] GU J, HAN Z, CHEN S, et al. A systematic survey of prompt engineering on vision-language foundation models[J]. arXiv:2307.12980, 2023.
[9] PETERS M E, NEUMANN M, IYYER M, et al. Deep contextualized word representations[J]. arXiv:1802.05365, 2018.
[10] RADFORD A, NARASIMHAN K, SALIMANS T, et al. Improving language understanding by generative pre-training[EB/OL]. (2020-09-26)[2023-09-26]. https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf.
[11] DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[J]. arXiv:1810.04805, 2018.
[12] FU Z, YANG H, SO A M C, et al. On the effectiveness of parameter-efficient fine-tuning[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2023: 12799-12807.
[13] HAN X, ZHANG Z, DING N, et al. Pre-trained models: past, present and future[J]. AI Open, 2021, 2: 225-250.
[14] 陈德光, 马金林, 马自萍, 等. 自然语言处理预训练技术综述[J]. 计算机科学与探索, 2021, 15(8): 1359-1389.
CHEN G D, MA J L, MA Z P, et al. Review of pre-training techniques for natural language processing[J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(8): 1359-1389.
[15] ZHANG Z, WANG B. Prompt learning for news recommendation[J]. arXiv:2304.05263, 2023.
[16] CHANG K W, TSENG W C, LI S W, et al. Speechprompt: an exploration of prompt tuning on generative spoken language model for speech processing tasks[J]. arXiv:2203.
16773, 2022.
[17] LIU P, YUAN W, FU J, et al. Pre-train, prompt, and predict: a systematic survey of prompting methods in natural language processing[J]. ACM Computing Surveys, 2023, 55(9): 1-35.
[18] PETRONI F, ROCKT?SCHEL T, LEWIS P, et al. Language models as knowledge bases?[J]. arXiv:1909.01066, 2019.
[19] BROWN T, MANN B, RYDER N, et al. Language models are few-shot learners[C]//Advances in Neural Information Processing Systems, 2020: 1877-1901.
[20] SCHICK T, SCHüTZE H. Exploiting cloze questions for few shot text classification and natural language inference[J]. arXiv:2001.07676, 2020.
[21] JIANG Z, XU F F, ARAKI J, et al. How can we know what language models know?[J]. Transactions of the Association for Computational Linguistics, 2020, 8: 423-438.
[22] YUAN W, NEUBIG G, LIU P. BARTscore: evaluating generated text as text generation[C]//Advances in Neural Information Processing Systems, 2021: 27263-27277.
[23] HAVIV A, BERANT J, GLOBERSON A. BERTese: learning to speak to BERT[J]. arXiv:2103.05327, 2021.
[24] WALLACE E, FENG S, KANDPAL N, et al. Universal adversarial triggers for attacking and analyzing nlp[J]. arXiv:1908.07125, 2019.
[25] SHIN T, RAZEGHI Y, LOGAN IV R L, et al. AutoPrompt: eliciting knowledge from language models with automatically generated prompts[J]. arXiv:2010.15980, 2020.
[26] GAO T, FISCH A, CHEN D. Making pre-trained language models better few-shot learners[J]. arXiv:2012.15723, 2020.
[27] DAVISON J, FELDMAN J, RUSH A M. Commonsense knowledge mining from pretrained models[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, 2019: 1173-1178.
[28] LI X L, LIANG P. Warp-tuning: optimizing continuous prompts for generation[J]. arXiv:2101.00190, 2021.
[29] LESTER B, AL-RFOU R, CONSTANT N. The power of scale for parameter-efficient prompt tuning[J]. arXiv:2104.
08691, 2021.
[30] ZHONG Z, FRIEDMAN D, CHEN D. Factual probing is [mask]: learning vs. learning to recall[J]. arXiv:2104.05240, 2021.
[31] QIN G, EISNER J. Learning how to ask: querying lms with mixtures of soft prompts[J]. arXiv:2104.06599, 2021.
[32] HAMBARDZUMYAN K, KHACHATRIAN H, MAY J. Warp: word-level adversarial reprogramming[J]. arXiv:2101.
00121, 2021.
[33] LIU X, ZHENG Y, DU Z, et al. GPT understands, too[J]. arXiv:2103.10385, 2021.
[34] LIU X, JI K, FU Y, et al. P-tuning v2: prompt tuning can be comparable to fine-tuning universally across scales and tasks[J]. arXiv:2110.07602, 2021.
[35] HAN X, ZHAO W, DING N, et al. PTR: prompt tuning with rules for text classification[J]. AI Open, 2022, 3: 182-192.
[36] ZHANG R, SUN Y, YANG J, et al. Knowledge-augmented frame semantic parsing with hybrid prompt-tuning[C]//Proceedings of the International Conference on Acoustics, Speech and Signal Processing, 2023: 1-5.
[37] PEREZ E, KIELA D, CHO K. True few-shot learning with language models[C]//Advances in Neural Information Processing Systems, 2021: 11054-11070.
[38] AGHAJANYAN A, OKHONKO D, LEWIS M, et al. HTLM: hyper-text pre-training and prompting of language models[J]. arXiv:2107.06955, 2021.
[39] ZHAO M, SCHüTZE H. Discrete and soft prompting for multilingual models[J]. arXiv:2109.03630, 2021.
[40] GU Y, HAN X, LIU Z, et al. PPT: pre-trained prompt tuning for few-shot learning[J]. arXiv:2109.04332, 2021.
[41] 鲍琛龙, 吕明阳, 唐晋韬, 等. 与知识相结合的提示学习研究综述[J]. 中文信息学报, 2023, 37(7): 1-12.
BAO C L, LYU M Y, TANG J T, et al. A survey of prompt learning combined with knowledge[J]. Journal of Chinese Information Processing, 2023, 37(7): 1-12.
[42] HU S, DING N, WANG H, et al. Knowledgeable prompt-tuning: incorporating knowledge into prompt verbalizer for text classification[J]. arXiv:2108.02035, 2021.
[43] WANG J, WANG C, LUO F, et al. Towards unified prompt tuning for few-shot text classification[J]. arXiv:2205.05313, 2022.
[44] ZHU Y, ZHOU X, QIANG J, et al. Prompt-learning for short text classification[J]. arXiv:2202.11345, 2022.
[45] LI J, TANG T, NIE J Y, et al. Learning to transfer prompts for text generation[J]. arXiv:2205.01543, 2022.
[46] ZHANG L, GAO X. Transfer adaptation learning: a decade survey[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(1): 23-44.
[47] VU T, LESTER B, CONSTANT N, et al. Spot: better frozen model adaptation through soft prompt transfer[J]. arXiv:2110.07904, 2021.
[48] WANG C, WANG J, QIU M, et al. Transprompt: towards an automatic transferable prompting framework for few-shot text classification[C]//Proceedings of the 2021 Conference on Empirical Methods In Natural Language Processing, 2021: 2792-2802.
[49] WANG S, FANG H, KHABSA M, et al. Entailment as few-shot learner[J]. arXiv:2104.14690, 2021.
[50] SUN Y, ZHENG Y, HAO C, et al. NSP-BERT: a prompt-based few-shot learner through an original pre-training task-next sentence prediction[J]. arXiv:2109.03564, 2021.
[51] HARRANDO I, REBOUD A, SCHLEIDER T, et al. ProZe: explainable and prompt-guided zero-shot text classification[J]. IEEE Internet Computing, 2022, 26(6): 69-77.
[52] DAN Y, ZHOU J, CHEN Q, et al. Enhancing class understanding via prompt-tuning for zero-shot text classification[C]//Proceedings of the International Conference on Acoustics, Speech and Signal Processing, 2022: 4303-4307.
[53] GUO Y, CODELLA N C, KARLINSKY L, et al. A broader study of cross-domain few-shot learning[C]//Proceedings of the 16th European Conference on Computer Vision, Glasgow, UK, August 23-28, 2020: 124-141.
[54] MAYER C W F, LUDWIG S, BRANDT S. Prompt text classifications with transformer models! an exemplary introduction to prompt-based learning with large language models[J]. Journal of Research on Technology in Education, 2023, 55(1): 125-141.
[55] SONG C, SHAO T, LIN K, et al. Investigating prompt learning for Chinese few-shot text classification with pre-trained language models[J]. Applied Sciences, 2022, 12(21): 11117.
[56] WANG H, XU C, MCAULEY J. Automatic multi-label prompting: simple and interpretable few-shot classification[J]. arXiv:2204.06305, 2022.
[57] 于碧辉, 蔡兴业, 魏靖烜. 基于提示学习的小样本文本分类方法[J]. 计算机应用, 2023, 43(9): 2735-2740.
YU B H, CAI X Y, WEI J X. Few-shot text classification method based on prompt learning[J]. Journal of Computer Applications, 2023, 43(9): 2735-2740.
[58] HU T, CHEN Z, GE J, et al. A Chinese few-shot text classification method utilizing improved prompt learning and unlabeled data[J]. Applied Sciences, 2023, 13(5): 3334.
[59] WEI L, LI Y, ZHU Y, et al. Prompt tuning for multi-label text classification: how to link exercises to knowledge concepts?[J]. Applied Sciences, 2022, 12(20): 10363.
[60] SONG R, LIU Z, CHEN X, et al. Label prompt for multi-label text classification[J]. Applied Intelligence, 2023, 53(8): 8761-8775.