基于关键句和题型的阅读理解问题生成技术研究

doi:10.3778/j.issn.1002-8331.2403-0231

摘要/Abstract

摘要： 现有的阅读理解问题生成技术研究主要以答案已知为前提展开。为了使阅读理解问题生成技术可以摆脱对答案的依赖，从而促使教育领域的问题生成技术向着端到端自动化出题的应用场景进一步迈进，对答案未知前提下的阅读理解问题生成技术展开了研究。提出引入关键句和题型作为两种简单的控制信息，替代需要从文章中抽取总结的答案，规定问题的提问内容范围和提问特点。提出了一个基于关键句和题型信息的可控问题生成框架，通过给定文章、关键句和题型来生成阅读理解问题。通过自动化评测和人工评测，验证了框架包含的两种问题生成方法的有效性和先进性。框架所需的两种控制信息相比答案更容易获取，为问题生成技术的应用带来更好的易用性，为使用者提供更高的出题效率。

关键词: 英语阅读理解, 问题生成, 文本生成

Abstract: The existing reading comprehension question generation methods are mostly answer-aware, mainly based on the premise that the answers are known. To enhance the applicability of reading comprehension question generation techniques within the educational domain towards the goal of end-to-end automated question generation, this study investigates the reading comprehension question generation method under the premise that answers are unknown. This study adopts key sentences and question types as two simple control signals instead of extracting and summarizing answers from the passage to specify the scope and phrasing of the questions. This study proposes a controllable question generation framework to generate reading comprehension questions by providing certain passages, key sentences and question types. Both automated and human evaluations verify the effectiveness and advancement of the two question generation methods included in the framework. The control signals required by the framework is easier to obtain than answers, which improves user experience of question generation techniques and provides users with higher question generation efficiency.

Key words: English reading comprehension, question generation, text generation

蒋玉茹, 陶宇阳, 王霞, 葛诗利. 基于关键句和题型的阅读理解问题生成技术研究[J]. 计算机工程与应用, 2025, 61(12): 141-153.

JIANG Yuru, TAO Yuyang, WANG Xia, GE Shili. Research on Reading Comprehension Question Generation Technique Based on Key Sentences and Question Types[J]. Computer Engineering and Applications, 2025, 61(12): 141-153.

参考文献

[1] 杨生文. 基于深度学习的阅读理解题目生成研究[D]. 武汉: 华中师范大学, 2021.
YANG S W. Research on generation of reading comprehension subjects based on deep learning[D]. Wuhan: Central China Normal University, 2021.
[2] 帅鹏举. 基于序列到序列的问题及干扰项生成方法研究[D]. 重庆: 西南大学, 2022.
SHUAI P J. Research on sequence-to-sequence based questionand distractor generation methods[D]. Chongqing: Southwest University, 2022.
[3] JIA X, ZHOU W, SUN X, et al. EQG-RACE: examination-type question generation[C]//Proceedings of the 35th AAAI Conference on Artificial Intelligence (AAAI 2021), 2021: 13143-13151.
[4] PAN Y C, HU B T, WANG S Y, et al. Learning to generate complex question with intent prediction from long passage[J]. Applied Intelligence, 2023, 53(5): 5823-5833.
[5] XU J, SUN Y, GAN J H, et al. Leveraging structured information from a passage to generate questions[J]. Tsinghua Science and Technology, 2023, 28(3): 464-474.
[6] LAI G K, XIE Q Z, LIU H X, et al. RACE: large-scale reading comprehension dataset from examinations[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2017: 785-794.
[7] ZYRIANOVA M, KALPAKCHI D, BOYE J. EMBRACE: evaluation and modifications for boosting RACE[J]. arXiv:2305.08433, 2023.
[8] RAJPURKAR P, ZHANG J, LOPYREV K, et al. SQuAD: 100,000+ questions for machine comprehension of text[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2016: 2383-2392.
[9] YANG Z L, QI P, ZHANG S Z, et al. HotpotQA: a dataset for diverse, explainable multi-hop question answering[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2018: 2369-2380.
[10] TRISCHLER A, WANG T, YUAN X D, et al. NewsQA: a machine comprehension dataset[C]//Proceedings of the 2nd Workshop on Representation Learning for NLP. Stroudsburg: ACL, 2017: 191-200.
[11] LIANG Y, LI J, YIN J. A new multi-choice reading comprehension dataset for curriculum learning[C]//Proceedings of the Asian Conference on Machine Learning, 2019: 742-757.
[12] YU W H, JIANG Z H, DONG Y F, et al. ReClor: a reading comprehension dataset requiring logical reasoning[J]. arXiv:2002.04326, 2020.
[13] WELBL J, LIU N F, GARDNER M. Crowdsourcing multiple choice science questions[C]//Proceedings of the 3rd Workshop on Noisy User-generated Text. Stroudsburg: ACL, 2017: 94-106.
[14] JAUHAR S K, TURNEY P, HOVY E. TabMCQ: a dataset of general knowledge tables and multiple-choice questions[J]. arXiv:1602.03960, 2016.
[15] HADIFAR A, BITEW S K, DELEU J, et al. EduQG: a multi-format multiple-choice dataset for the educational domain[J]. IEEE Access, 2023, 11: 20885-20896.
[16] DU X Y, CARDIE C. Identifying where to focus in reading comprehension for NeuralQuestion generation[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2017: 2067-2073.
[17] DUAN N, TANG D Y, CHEN P, et al. Question generation for question answering[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2017: 866-874.
[18] SUN X W, LIU J, LYU Y J, et al. Answer-focused and position-aware neural question generation[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2018: 3930-3939.
[19] GAO Y F, BING L D, LI P J, et al. Generating distractors for reading comprehension questions from real examinations[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2019: 6423-6430.
[20] SONG L F, WANG Z G, HAMZA W, et al. Leveraging context information for natural question generation[C]//Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2. Stroudsburg: ACL, 2018: 569-574.
[21] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[22] DEVLIN J, CHANG M, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language under-standing[C]///Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), 2019: 4171-4186.
[23] KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[C]//Proceedings of the International Conference on Learning Representations, 2016.
[24] VELI?KOVI? P, CUCURULL G, CASANOVA A, et al. Graph Attention Networks[C]//Proceedings of the International Conference on Learning Representations, 2018.
[25] SHUAI P J, LI L, LIU S S, et al. QDG: a unified model for automatic question-distractor pairs generation[J]. Applied Intelligence, 2023, 53(7): 8275-8285.
[26] SACHAN M, XING E. Self-training for jointly learning to ask and answer questions[C]//Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1. Stroudsburg: ACL, 2018: 629-640.
[27] DU X Y, CARDIE C. Harvesting paragraph-level question-answer pairs from wikipedia[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2018: 1907-1917.
[28] WILLIS A, DAVIS G, RUAN S, et al. Key phrase extraction for generating educational question-answer pairs[C]//Proceedings of the Sixth (2019) ACM Conference on Learning @ Scale. New York: ACM, 2019: 1-10.
[29] QU F Y, JIA X, WU Y F. Asking questions like educational experts: automatically generating question-answer pairs on real-world examination data[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. Stroudsburg: ACL, 2021: 2583-2593.
[30] QI W Z, YAN Y, GONG Y Y, et al. ProphetNet: predicting future N-gram for sequence-to-sequence pre-training[C]//Findings of the Association for Computational Linguistics: EMNLP 2020. Stroudsburg: ACL, 2020: 2401-2410.
[31] GONZALEZ H, DUGAN L, MILTSAKAKI E, et al. Enhancing human summaries for question-answer generation in education[C]//Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications. Stroudsburg: ACL, 2023: 108-118.
[32] COLIN R, NOAM S, ADAM R, et al. Exploring the limits of transfer learning with a unified text-to-text transformer[J]. Journal of Machine Learning Research, 2020, 21(1): 5485?-?5551.
[33] BROWN T, MANN B, RYDER N, et al. Language models are few-shot learners[C]//Advances in Neural Information Processing Systems, 2020: 1877-1901.
[34] WANG Y C, LI L. Generating question-answer pairs for few-shot learning[C]//Proceedings of the International Conference on Artificial Neural Networks, 2023: 414-425.
[35] ACHIAM J, ADLER S, AGARWAL S, et al. GPT-4 technical report[J]. arXiv:2303.08774, 2023.
[36] REIMERS N, GUREVYCH I. Sentence-BERT: sentence embeddings using Siamese BERT-networks[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing. Stroudsburg: ACL, 2019: 3980-3990.
[37] CHEN D Q, BOLTON J, MANNING C D. A thorough examination of the CNN/daily mail reading comprehension task[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2016: 2358-2367.
[38] BIRD S, KLEIN E, LOPER E. Natural language processing with Python: analyzing text with the natural language toolkit[M]. [S.l.]: O’Reilly Media, Inc., 2009.
[39] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Pro-Cessing Systems, 2017.
[40] ZHANG T, KISHORE V, WU F, et al. BERTScore: evaluating text generation with BERT[C]//Proceedings of the International Conference on Learning Representations, 2019.
[41] HOSKING T, RIEDEL S. Evaluating rewards for question generation models[C]//Proceedings of the 2019 Conference of the North. Stroudsburg: ACL, 2019: 2278-2283.
[42] SEE A, LIU P J, MANNING C D. Get to the point: summarization with pointer-generator networks[C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2017: 1073-1083.
[43] ZHANG S Y, BANSAL M. Addressing semantic drift in question generation for semi-supervised question answering[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing. Stroudsburg: ACL, 2019: 2495-2509.
[44] LEWIS M, LIU Y H, GOYAL N, et al. BART: denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2020: 7871-7880.
[45] CHUNG H W, HOU L, LONGPRE S, et al. Scaling instruction-finetuned language models[J]. arXiv:2210.11416, 2022.