Journal of Computer Applications ›› 2022, Vol. 42 ›› Issue (3): 867-873.DOI: 10.11772/j.issn.1001-9081.2021030375
Special Issue: 人工智能
• Artificial intelligence • Previous Articles Next Articles
Zengzhen DU1, Dongxin TANG2, Dan XIE1()
Received:
2021-03-15
Revised:
2021-05-20
Accepted:
2021-05-31
Online:
2022-04-09
Published:
2022-03-10
Contact:
Dan XIE
About author:
DU Zengzhen, born in 1998, M. S. candidate. Her research interests include natural language processing, artificial intelligence.Supported by:
通讯作者:
解丹
作者简介:
杜曾贞(1998—),女,湖北荆州人,硕士研究生,主要研究方向:自然语言处理、人工智能基金资助:
CLC Number:
Zengzhen DU, Dongxin TANG, Dan XIE. Method of generating rhetorical questions based on deep neural network in intelligent consultation[J]. Journal of Computer Applications, 2022, 42(3): 867-873.
杜曾贞, 唐东昕, 解丹. 智能问诊中基于深度神经网络的反问生成方法[J]. 《计算机应用》唯一官方网站, 2022, 42(3): 867-873.
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.joca.cn/EN/10.11772/j.issn.1001-9081.2021030375
角色 | 对话 |
---|---|
患者 | 弱视属于视力明显下降吗? |
医生 | 弱视是指视力要比健康眼差。 |
患者 | 我的孩子检查出弱视。 |
医生 | 孩子是两个眼睛视力不好还是一只呢? |
患者 | 一只眼睛。 |
医生 | 方便提供孩子的裸眼和矫正视力吗? |
Tab.1 Question-answer example between a doctor and a patient
角色 | 对话 |
---|---|
患者 | 弱视属于视力明显下降吗? |
医生 | 弱视是指视力要比健康眼差。 |
患者 | 我的孩子检查出弱视。 |
医生 | 孩子是两个眼睛视力不好还是一只呢? |
患者 | 一只眼睛。 |
医生 | 方便提供孩子的裸眼和矫正视力吗? |
数据集数量 | TextCNN | TextRNN | 数据集数量 | TextCNN | TextRNN |
---|---|---|---|---|---|
1 000 | 84.60 | 92.50 | 10 000 | 89.50 | 96.25 |
2 000 | 86.50 | 94.00 | 20 000 | 89.00 | 96.00 |
5 000 | 88.00 | 95.50 | 30 000 | 88.65 | 95.75 |
Tab. 2 Accuracy test results of different selected datasets
数据集数量 | TextCNN | TextRNN | 数据集数量 | TextCNN | TextRNN |
---|---|---|---|---|---|
1 000 | 84.60 | 92.50 | 10 000 | 89.50 | 96.25 |
2 000 | 86.50 | 94.00 | 20 000 | 89.00 | 96.00 |
5 000 | 88.00 | 95.50 | 30 000 | 88.65 | 95.75 |
选取方式 | BLSTM-QuD | BERT-QuD | |||
---|---|---|---|---|---|
F1-macro | F1-micro | Accuracy | F1 | Accuracy | |
Window-top | 0.715 1 | 0.715 7 | 0.663 1 | 0.696 3 | 0.678 7 |
Window-last-q | 0.742 3 | 0.745 6 | 0.692 8 | 0.753 8 | 0.749 6 |
Window-1 | 0.704 1 | 0.704 1 | 0.667 1 | 0.705 4 | 0.709 0 |
Window-3 | 0.702 8 | 0.703 1 | 0.659 6 | 0.718 7 | 0.712 2 |
Window-5 | 0.696 3 | 0.696 4 | 0.644 8 | 0.725 3 | 0.726 0 |
Window-10 | 0.677 6 | 0.677 7 | 0.610 7 | 0.671 4 | 0.669 2 |
Tab. 3 Comparison of question trigger detection results based on six selection methods
选取方式 | BLSTM-QuD | BERT-QuD | |||
---|---|---|---|---|---|
F1-macro | F1-micro | Accuracy | F1 | Accuracy | |
Window-top | 0.715 1 | 0.715 7 | 0.663 1 | 0.696 3 | 0.678 7 |
Window-last-q | 0.742 3 | 0.745 6 | 0.692 8 | 0.753 8 | 0.749 6 |
Window-1 | 0.704 1 | 0.704 1 | 0.667 1 | 0.705 4 | 0.709 0 |
Window-3 | 0.702 8 | 0.703 1 | 0.659 6 | 0.718 7 | 0.712 2 |
Window-5 | 0.696 3 | 0.696 4 | 0.644 8 | 0.725 3 | 0.726 0 |
Window-10 | 0.677 6 | 0.677 7 | 0.610 7 | 0.671 4 | 0.669 2 |
选取方式 | 机器评估 | 人工评价/% | |||||||
---|---|---|---|---|---|---|---|---|---|
BLEU-1 | BLEU-2 | BLEU-4 | PPL | “0” | “1” | “2” | “3” | Average | |
Window-top | 0.360 0 | 0.254 5 | 0.111 7 | 1.569 7 | 0.40 | 19.20 | 55.60 | 24.80 | 51.20 |
Window-last-q | 0.310 4 | 0.209 0 | 0.053 6 | 1.864 3 | 0.80 | 22.60 | 49.80 | 26.80 | 50.65 |
Window-1 | 0.062 0 | 0.007 9 | 0.001 8 | 1.824 8 | 3.60 | 48.60 | 35.80 | 12.00 | 39.05 |
Window-3 | 0.287 5 | 0.204 8 | 0.057 0 | 1.885 4 | 0.60 | 31.00 | 45.40 | 23.00 | 47.70 |
Window-5 | 0.326 2 | 0.226 0 | 0.056 5 | 1.973 7 | 1.40 | 34.00 | 43.20 | 21.40 | 46.15 |
Window-10 | 0.305 3 | 0.212 2 | 0.060 9 | 1.968 6 | 4.20 | 37.40 | 41.40 | 17.00 | 42.80 |
Tab. 4 Comparison of rhetorical question generation results based on six selection methods
选取方式 | 机器评估 | 人工评价/% | |||||||
---|---|---|---|---|---|---|---|---|---|
BLEU-1 | BLEU-2 | BLEU-4 | PPL | “0” | “1” | “2” | “3” | Average | |
Window-top | 0.360 0 | 0.254 5 | 0.111 7 | 1.569 7 | 0.40 | 19.20 | 55.60 | 24.80 | 51.20 |
Window-last-q | 0.310 4 | 0.209 0 | 0.053 6 | 1.864 3 | 0.80 | 22.60 | 49.80 | 26.80 | 50.65 |
Window-1 | 0.062 0 | 0.007 9 | 0.001 8 | 1.824 8 | 3.60 | 48.60 | 35.80 | 12.00 | 39.05 |
Window-3 | 0.287 5 | 0.204 8 | 0.057 0 | 1.885 4 | 0.60 | 31.00 | 45.40 | 23.00 | 47.70 |
Window-5 | 0.326 2 | 0.226 0 | 0.056 5 | 1.973 7 | 1.40 | 34.00 | 43.20 | 21.40 | 46.15 |
Window-10 | 0.305 3 | 0.212 2 | 0.060 9 | 1.968 6 | 4.20 | 37.40 | 41.40 | 17.00 | 42.80 |
类别 | 实例1 | 实例2 |
---|---|---|
上文 | 患者:慢性鼻炎怎么治疗啊? 医生:您好。平时是什么症状呢? 患者:鼻子不通气。 | 患者:弱视属于视力明显下降吗? 医生:弱视是指视力要比健康眼差。 患者:我的孩子检查出是弱视。 医生:孩子是两个眼睛视力不好还是一只呢? 患者:一只眼睛。 |
原文 | 医生:持续多久了? | 医生:方便提供孩子的裸眼和矫正视力吗? |
Window-top | 医生:您好,这种情况多久了? | 医生:已经持续一阵子了吗? |
Window-last-q | 医生,您好,做过什么检查? | 医生:视力是多少呢? |
Tab. 5 Two examples of rhetorical question generation between a doctor and a patient
类别 | 实例1 | 实例2 |
---|---|---|
上文 | 患者:慢性鼻炎怎么治疗啊? 医生:您好。平时是什么症状呢? 患者:鼻子不通气。 | 患者:弱视属于视力明显下降吗? 医生:弱视是指视力要比健康眼差。 患者:我的孩子检查出是弱视。 医生:孩子是两个眼睛视力不好还是一只呢? 患者:一只眼睛。 |
原文 | 医生:持续多久了? | 医生:方便提供孩子的裸眼和矫正视力吗? |
Window-top | 医生:您好,这种情况多久了? | 医生:已经持续一阵子了吗? |
Window-last-q | 医生,您好,做过什么检查? | 医生:视力是多少呢? |
得分 | 对话长度<5 | 对话长度≥5 |
---|---|---|
1 | 75.2 | 30.6 |
0 | 24.8 | 69.4 |
Tab.6 Score proportion of short context and long context
得分 | 对话长度<5 | 对话长度≥5 |
---|---|---|
1 | 75.2 | 30.6 |
0 | 24.8 | 69.4 |
1 | EASTIN C, EASTIN T. Clinical characteristics of coronavirus disease 2019 in China[J]. Journal of Emergency Medicine, 2020, 58(4): 711-712. 10.1016/j.jemermed.2020.04.004 |
2 | 王智悦,于清,王楠,等. 基于知识图谱的智能问答研究综述[J]. 计算机工程与应用, 2020, 56(23): 1-11. 10.1109/aeeca52519.2021.9574313 |
WANG Z Y, YU Q, WANG N, et al. Survey of intelligent question answering research based on knowledge graph[J]. Computer Engineering and Applications, 2020, 56(23): 1-11. 10.1109/aeeca52519.2021.9574313 | |
3 | 贺佳,杜建强,聂斌,等. 智能问答系统在医学领域的应用研究[J]. 医学信息, 2018, 31(14): 16-19. 10.3969/j.issn.1006-1959.2018.14.007 |
HE J, DU J Q, NIE B, et al. Research on the application of intelligent question-answering system in medical field[J]. Medical Information, 2018, 31(14): 16-19. 10.3969/j.issn.1006-1959.2018.14.007 | |
4 | LEE M, CIMINO J, HAI R Z, et al. Beyond information retrieval-medical question answering[J]. American Medical Informatics Association Annual Symposium Proceedings, 2006, 1(1): 469-473. |
5 | CAO Y G, LIU F, SIMPSON P, et al. AskHERMES: an online question answering system for complex clinical questions[J]. Journal of Biomedical Informatics, 2011, 44(2): 277-288. 10.1016/j.jbi.2011.01.004 |
6 | ABACHA A BEN, ZWEIGENBAUM P. MEANS: a medical question-answering system combining NLP techniques and semantic Web technologies[J]. Information Processing and Management, 2015, 51(5): 570-594. 10.1016/j.ipm.2015.04.006 |
7 | ASIAEE A H, MINNING T, DOSHI P, et al. A framework for ontology-based question answering with application to parasite immunology[J]. Journal of Biomedical Semantics, 2015(6): Article No. 31. 10.1186/s13326-015-0029-x |
8 | YIN Y, ZHANG Y, LIU X, et al. HealthQA: a Chinese QA summary system for smart health[C]// Proceedings of the 2014 International Conference on Smart Health. Cham: Springer, 2014:51-62. 10.1007/978-3-319-08416-9_6 |
9 | 田迎,单娅辉,王时绘. 基于知识图谱的抑郁症自动问答系统研究[J]. 湖北大学学报(自然科学版), 2020, 42(5): 587-591. |
TIAN Y, SHAN Y H, WANG S H. The research of depression automatic question answering system based on knowledge graph[J]. Journal of Hubei University (Natural Science), 2020, 42(5): 587-591. | |
10 | 乔宇,崔亮亮,李帅,等. 智能问答机器人系统研发及应用研究:以济南市新型冠状病毒肺炎疫情处置应对为例[J]. 山东大学学报(医学版), 2020, 58(4): 17-22. |
QIAO Y, CUI L L, LI S, et al. Research and development of intelligent question answering robot system: a case study of its application in response to COVID-19 epidemic in Jinan City[J]. Journal of Shandong University (Health Sciences), 2020, 58(4): 17-22. | |
11 | 汤人杰,杨巧节. 基于医疗知识图谱的智能辅助问诊模型研究[J]. 中国数字医学, 2020, 15(10): 5-8. 10.3969/j.issn.1673-7571.2020.10.002 |
TANG R J, YANG Q J. Research on the intelligent assisted inquiry model based on medical knowledge map[J]. China Digital Medicine, 2020, 15(10): 5-8. 10.3969/j.issn.1673-7571.2020.10.002 | |
12 | 王浩. 神经机器翻译模型的实现验证及其剪枝压缩[D]. 南京:南京大学, 2019:2.(WANG H. Implementation, verification and compression by pruning of neural machine translation model[D]. Nanjing: Nanjing University, 2019:2.) |
13 | WU Y, SCHUSTER M, CHEN Z, et al. Google’s neural machine translation system: bridging the gap between human and machine translation[EB/OL].[2020-06-22]. . |
14 | OTT M, EDUNOV S, BAEVSKI A, et al. Fairseq: a fast, extensible toolkit for sequence modeling[EB/OL].[2020-06-22]. . 10.18653/v1/n19-4009 |
15 | KLEIN G, KIM Y, DENG Y T, et al. OpenNMT: open-source toolkit for neural machine translation[C]// Proceedings of the 2017 Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics, 2017:67-72. 10.18653/v1/p17-4012 |
16 | 周海林,沈志贤. 谷歌神经机器翻译质量现状分析[J]. 科技资讯, 2018, 16(1): 220. 10.16661/j.cnki.1672-3791.2018.01.220 |
ZHOU H L, SHEN Z X. Analysis of the quality of Google’s neural machine translation[J]. Science and Technology Information, 2018, 16(1): 220. 10.16661/j.cnki.1672-3791.2018.01.220 | |
17 | 牛向华,苏依拉,高芬,等. 单语数据训练在蒙汉神经机器翻译中的应用[J]. 计算机应用与软件, 2020, 37(6):178-183. 10.3969/j.issn.1000-386x.2020.06.032 |
NIU X H, SU Y L, GAO F, et al. Application of monolingual data training on Mongolian-Chinese neural machine translation[J]. Computer Applications and Software, 2020, 37(6):178-183. 10.3969/j.issn.1000-386x.2020.06.032 | |
18 | KIM Y. Convolutional neural networks for sentence classification[C]// Proceedings of the 2014 Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics, 2014:1746-1751. 10.3115/v1/d14-1181 |
19 | 李雅昆,潘晴, WANG E X.基于改进的多层BLSTM的中文分词和标点预测[J]. 计算机应用, 2018, 38(5): 1278-1282, 1314. 10.11772/j.issn.1001-9081.201711112631 |
LI Y K, PAN Q, WANG F. Joint Chinese word segmentation and punctuation prediction based on improved multilayer BLSTM network[J]. Journal of Computer Applications, 2018, 38(5): 1278-1282, 1314. 10.11772/j.issn.1001-9081.201711112631 | |
20 | 华冰涛,袁志祥,肖维民,等. 基于BLSTM-CNN-CRF模型的槽填充与意图识别[J]. 计算机工程与应用, 2019, 55(9): 139-143. 10.3778/j.issn.1002-8331.1801-0183 |
HUA B T, YUAN Z X, XIAO W M, et al. Joint slot filling and intent detection with BLSTM-CNN-CRF[J]. Computer Engineering and Applications, 2019, 55(9): 139-143. 10.3778/j.issn.1002-8331.1801-0183 | |
21 | 严佩敏,唐婉琪. 基于改进BERT的中文文本分类[J]. 工业控制计算机, 2020, 33(7): 108-110. 10.3969/j.issn.1001-182X.2020.07.041 |
YAN P M, TANG W Q. Chinese text classification based on improved BERT[J]. Industrial Control Computer, 2020, 33(7): 108-110. 10.3969/j.issn.1001-182X.2020.07.041 | |
22 | 张东东,彭敦陆. ENT-BERT:结合BERT和实体信息的实体关系分类模型[J]. 小型微型计算机系统, 2020, 41(12): 2557-2562. 10.3969/j.issn.1000-1220.2020.12.016 |
ZHANG D D, PENG D L. ENT-BERT: entity relation classification model combining BERT and entity information[J]. Journal of Chinese Computer Systems, 2020, 41(12): 2257-2562. 10.3969/j.issn.1000-1220.2020.12.016 | |
23 | 张小川,戴旭尧,刘璐,等. 融合多头自注意力机制的中文短文本分类模型[J]. 计算机应用, 2020, 40(12): 3485-3489. 10.11772/j.issn.1001-9081.2020060914 |
ZHANG X C, DAI X Y, LIU L, et al. Chinese short text classification model with multi-head self-attention mechanism[J]. Journal of Computer Applications, 2020, 40(12): 3485-3489. 10.11772/j.issn.1001-9081.2020060914 | |
24 | 王月,王孟轩,张胜,等. 基于BERT的警情文本命名实体识别[J]. 计算机应用, 2020, 40(2): 535-540. 10.11772/j.issn.1001-9081.2019101717 |
WANG Y, WANG M X, ZHANG S, et al. Alarm text named entity recognition based on BERT[J]. Journal of Computer Applications, 2020, 40(2): 535-540. 10.11772/j.issn.1001-9081.2019101717 | |
25 | LUO W, WU D. Research and implementation of Seq2Seq model chat robot based on attention mechanism[J]. Journal of Physics: Conference Series, 2020, 1693(1): 1-8. 10.1088/1742-6596/1693/1/012200 |
26 | QIU L, LI J, BI W, et al. Are training samples correlated? learning to generate dialogue responses with multiple references[C]// Proceedings of the 2019 Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics, 2019:3826-3835. 10.18653/v1/p19-1372 |
27 | SERBAN I V, SORDONI A, LOWE R, et al. A hierarchical latent variable encoder-decoder model for generating dialogues[C]// Proceedings of the 2017 AAAI Conference on Artificial Intelligence. Palo Alto, CA: AAAI Press, 2017:3295-3301. |
28 | LI J W, MONROE W, SHI T L, et al. Adversarial learning for neural dialogue generation[C]// Proceedings of the 2017 Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics, 2017:2157-2169. |
29 | 乔博文,李军辉. 融合语义角色的神经机器翻译[J]. 计算机科学, 2020, 47(2): 163-168. 10.11896/jsjkx.190100048 |
QIAO B W, LI J H. Neural machine translation combining source semantic roles[J]. Computer Science, 2020, 47(2): 163-168. 10.11896/jsjkx.190100048 | |
30 | CHUANG S P, SUNG T W, LIU A H, et al. Worse WER, but better BLEU? leveraging word embedding as intermediate in multitask end-to-end speech translation[C]// Proceedings of the 2020 Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics, 2020:5998-6003. 10.18653/v1/2020.acl-main.533 |
31 | 陈俊航,徐小平,杨恒泓. 基于Seq2seq模型的推荐应用研究[J]. 计算机科学, 2019, 46(S1): 493-496. |
CHEN J H, XU X P, YANG H H. Research on recommendation application based on Seq2seq model[J]. Computer Science, 2019, 46(S1): 493-496. | |
32 | HE T W, WANG H, HU B F. Evaluate the Chinese version of machine translation based on perplexity analysis[C]// Proceedings of the 2017 International Conference on Computational Science and Engineering. Piscataway: IEEE, 2017:278-281. 10.1109/cse-euc.2017.57 |
[1] | Junfeng SHEN, Xingchen ZHOU, Can TANG. Dual-channel sentiment analysis model based on improved prompt learning method [J]. Journal of Computer Applications, 2024, 44(6): 1796-1806. |
[2] | Hua LAI, Tong SUN, Wenjun WANG, Zhengtao YU, Shengxiang GAO, Ling DONG. Text punctuation restoration for Vietnamese speech recognition with multimodal features [J]. Journal of Computer Applications, 2024, 44(2): 418-423. |
[3] | Ming XU, Linhao LI, Qiaoling QI, Liqin WANG. Abductive reasoning model based on attention balance list [J]. Journal of Computer Applications, 2023, 43(2): 349-355. |
[4] | Min ZUO, Hong WANG, Wenjing YAN, Qingchuan ZHANG. Gene splice site identification based on BERT and CNN [J]. Journal of Computer Applications, 2023, 43(10): 3309-3314. |
[5] | Xianfeng YANG, Jiahe ZHAO, Ziqiang LI. Text classification model combining word annotations [J]. Journal of Computer Applications, 2022, 42(5): 1317-1323. |
[6] | Haifeng ZHANG, Cheng ZENG, Lie PAN, Rusong HAO, Chaodong WEN, Peng HE. News topic text classification method based on BERT and feature projection network [J]. Journal of Computer Applications, 2022, 42(4): 1116-1124. |
[7] | Shoulong JIAO, Youxiang DUAN, Qifeng SUN, Zihao ZHUANG, Chenhao SUN. Knowledge representation learning method incorporating entity description information and neighbor node features [J]. Journal of Computer Applications, 2022, 42(4): 1050-1056. |
[8] | Xueqiang LYU, Chen PENG, Le ZHANG, Zhi’an DONG, Xindong YOU. Text multi-label classification method incorporating BERT and label semantic attention [J]. Journal of Computer Applications, 2022, 42(1): 57-63. |
[9] | CHEN Jiawei, HAN Fang, WANG Zhijie. Aspect-based sentiment analysis with self-attention gated graph convolutional network [J]. Journal of Computer Applications, 2020, 40(8): 2202-2206. |
[10] | YANG Guofeng, YANG Yong. Question classification of common crop disease question answering system based on BERT [J]. Journal of Computer Applications, 2020, 40(6): 1580-1586. |
Viewed | ||||||
Full text |
|
|||||
Abstract |
|
|||||