Recognition of sentencing circumstances in adjudication documents based on abductive learning

doi:10.11772/j.issn.1001-9081.2021091748

Journal of Computer Applications ›› 2022, Vol. 42 ›› Issue (6): 1802-1807.DOI: 10.11772/j.issn.1001-9081.2021091748

• The 18th CCF Conference on Web Information Systems and Applications • Previous Articles

Recognition of sentencing circumstances in adjudication documents based on abductive learning

Jinye LI¹, Ruizhang HUANG¹^,²(), Yongbin QIN¹^,², Yanping CHEN¹^,², Xiaoyu TIAN¹

^1.College of Computer Science and Technology，Guizhou University，Guiyang Guizhou 550025，China
^2.State Key Laboratory of Public Big Data （Guizhou University），Guiyang Guizhou 550025，China

Received:2021-10-12 Revised:2021-11-18 Accepted:2021-11-26 Online:2022-04-15 Published:2022-06-10
Contact: Ruizhang HUANG
About author:LI Jinye，born in 1997，M. S. candidate. His research interests include abductive learning.
QIN Yongbin，born in 1980，Ph. D.，professor. His research interests include intelligent computing，machine learning，algorithm design.
CHEN Yanping，born in 1980，Ph. D.，associate professor. His research interests include artificial intelligence，natural language processing.
TIAN Xiaoyu，born in 1997，M. S. candidate. Her research interests include abductive learning.
Supported by:
Natural Science Foundation of China(62066008);Key Project of Science and Technology Foundation of Guizhou Province （Qianke Hejichu ［2020］ 1Z055）

基于反绎学习的裁判文书量刑情节识别

李锦烨¹, 黄瑞章¹^,²(), 秦永彬¹^,², 陈艳平¹^,², 田小瑜¹

^1.贵州大学计算机科学与技术学院，贵阳 550025
^2.公共大数据国家重点实验室（贵州大学），贵阳 550025

通讯作者: 黄瑞章
作者简介:李锦烨（1997—），男，江苏泰州人，硕士研究生，主要研究方向：反绎学习
秦永彬（1980—），男，山东招远人，教授，博士，主要研究方向：智能计算、机器学习、算法设计
陈艳平（1980—），男，贵州长顺人，副教授，博士，CCF会员，主要研究方向：人工智能、自然语言处理
田小瑜（1997—），女，重庆人，硕士研究生，主要研究方向：反绎学习。
基金资助:
国家自然科学基金资助项目(62066008);贵州省科学技术基金重点项目(黔科合基础［2020］1Z055)

Abstract

Abstract:

Aiming at the problem of poor recognition of sentencing circumstances in adjudication documents caused by the lack of labeled data， low quality of labeling and existence of strong logicality in judicial field， a sentencing circumstance recognition model based on abductive learning named ABL-CON （ABductive Learning in CONfidence） was proposed. Firstly， combining with neural network and domain logic inference， through the semi-supervised method， a confidence learning method was used to characterize the confidence of circumstance recognition. Then， the illogical error circumstances generated by neural network of the unlabeled data were corrected， and the recognition model was retrained to improve the recognition accuracy. Experimental results on the self-constructed judicial dataset show that the ABL-CON model using 50% labeled data and 50% unlabeled data achieves 90.35% and 90.58% in Macro_F1 and Micro_F1， respectively， which is better than BERT （Bidirectional Encoder Representations from Transformers） and SS-ABL （Semi-Supervised ABductive Learning） under the same conditions， and also surpasses the BERT model using 100% labeled data. The ABL-CON model can effectively improve the logical rationality of labels as well as the recognition ability of labels by correcting illogical labels through logical abductive correctness.

Key words: sentencing circumstance recognition, semi-supervised learning, multi-label classification, abductive learning, confidence learning

摘要：

针对司法领域标记数据匮乏、标注质量不高、存在强逻辑性导致裁判文书量刑情节识别效果不佳的问题，提出一种基于反绎学习的量刑情节识别模型ABL-CON。首先结合神经网络与领域逻辑推理，通过半监督学习方法，使用置信学习方法表征情节识别置信度；然后修正无标签数据经过神经网络产生的不合逻辑的错误情节，重新训练识别模型，以提高识别精度。在自构建的司法数据集上的实验结果表明，使用50%标注数据与50%无标注数据的ABL-CON模型在Macro_F1值和Micro_F1值上分别达到了90.35%和90.58%，优于同样条件下的BERT和SS-ABL，也超越了使用100%标注数据的BERT模型。ABL-CON模型通过逻辑反绎修正不符合逻辑的标签能够有效提高标签的逻辑合理性以及标签的识别能力。

关键词: 量刑情节识别, 半监督学习, 多标签分类, 反绎学习, 置信学习

CLC Number:

TP391.1

Jinye LI, Ruizhang HUANG, Yongbin QIN, Yanping CHEN, Xiaoyu TIAN. Recognition of sentencing circumstances in adjudication documents based on abductive learning[J]. Journal of Computer Applications, 2022, 42(6): 1802-1807.

李锦烨, 黄瑞章, 秦永彬, 陈艳平, 田小瑜. 基于反绎学习的裁判文书量刑情节识别[J]. 《计算机应用》唯一官方网站, 2022, 42(6): 1802-1807.

Figures/Tables 8

Fig. 1 Example of structure of adjudication document

Fig. 2 Examples of Case Description Fragment

Tab. 1 Examples of description of sentencing circumstances in adjudication documents

案情描述	涉及量刑情节	情节说明
本案中，涉案赃物已发还被害人，被告人自愿认罪并取得被害人谅解，依法可从轻处罚	no_damage attitude forgive	被告人具有“退赔、认罪、取得谅解”情节
被告人刘X入户盗窃、为吸毒而盗窃，有前科一次，均应酌情从重处罚；被告人刘X能在庭审中自愿认罪，应依法从轻处罚	indoor attitude again	被告人具有“入户盗窃、认罪、累犯”情节

Tab. 2 Part of sentencing circumstances’ rules

逻辑描述

规则

说明

“入户”盗窃是指以非法占有为目的，进入他人生活的与外界相对隔离的住所实施盗窃的行为；“扒窃”是在公共场所或者公共交通工具上盗窃他人随身携带的财物

∃ i n d o o r ⋂ t h e f t → F a l s e

由于“入户”“扒窃”的场景没有交集，

在单次盗窃中，两种情节不共存

Tab. 2 Part of sentencing circumstances’ rules

逻辑描述

规则

说明

∃ i n d o o r ⋂ t h e f t → F a l s e

由于“入户”“扒窃”的场景没有交集，

在单次盗窃中，两种情节不共存

Fig. 3 Model framework of sentencing circumstance recognition in judgement document based on abductive learning

Tab. 3 Statistical information of dataset

量刑情节	类别	样本数
量刑情节	类别	训练集	测试集
合计		4 371	1 103
no_damage	退赔	689	207
attitude	认罪态度	1 283	234
surrender	自首	150	49
again	前科	1 138	166
young	未成年	167	88
forgive	谅解	221	53
tool	使用工具	192	66
indoor	入户	305	206
theft	扒窃	226	34

Tab. 4 Performance comparison of different models

模型	Macro_F1	Micro_F1
FastText-100	73.12	74.85
CNN-100	78.17	80.13
RNN-100	82.11	81.28
RCNN-100	80.31	80.48
BERT-100	89.84	88.89
BERT-50	88.13	88.15
SS-ABL-50	84.32	85.45
ABL-CON-50	90.35	90.58

Tab. 5 F1 values of single-category sentencing circumstances

模型	no_damage	attitude	surrender	again	young	forgive	tool	indoor	theft
FastText-100	65.61	86.08	77.27	86.88	63.24	91.59	64.29	61.92	61.22
CNN-100	85.09	85.71	73.02	89.69	74.58	82.11	64.56	69.57	79.25
RNN-100	75.92	86.98	90.32	88.34	73.68	90.57	70.06	75.43	87.72
RCNN-100	72.07	87.71	83.17	87.22	76.62	94.44	73.74	72.35	75.47
BERT-100	84.15	91.67	92.78	93.70	92.05	96.08	81.63	84.72	91.80
BERT-50	82.68	91.02	88.42	94.21	85.16	95.04	82.47	86.45	87.71
SS-ABL-50	85.01	89.71	80.41	92.26	84.74	96.15	65.83	81.40	83.33
ABL-CON-50	87.70	93.71	89.58	94.56	86.08	96.09	84.31	89.62	91.53

References 19

1	ZHOU Z H. Abductive learning： towards bridging machine learning and logical reasoning［J］. Science China Information Sciences， 2019， 62（7）： No.76101. 10.1007/s11432-018-9801-4
2	HUANG Y X， DAI W Z， YANG J， et al. Semi-supervised abductive learning and its application to theft judicial sentencing［C］// Proceedings of the 2020 IEEE International Conference on Data Mining. Piscataway： IEEE， 2020： 1070-1075. 10.1109/icdm50108.2020.00127
3	LIU C L， HSIEH C D. Exploring phrase-based classification of judicial documents for criminal charges in Chinese［C］// Proceedings of the 2006 International Symposium on Methodologies for Intelligent Systems， LNCS 4203/LNAI 4203. Berlin： Springer， 2006： 681-690.
4	KATZ D M， BOMMARITO M J， II， BLACKMAN J. A general approach for predicting the behavior of the Supreme Court of the United States［EB/OL］. （2017-01-17）［2021-05-22］.. 10.2139/ssrn.2463244
5	ZHONG H X， GUO Z P， TU C C， et al. Legal judgment prediction via topological learning［C］// Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Stroudsburg， PA： Association for Computational Linguistics， 2018： 3540-3549. 10.18653/v1/d18-1390
6	ANGELIDIS I， CHALKIDIS I， KOUBARAKIS M. Named entity recognition， linking and generation for Greek legislation［C］// Proceedings of the 31st International Conference on Legal Knowledge and Information Systems. Amsterdam： IOS Press， 2018： 1-10. 10.1145/3308560.3317077
7	CARDELLINO C， TERUEL M， ALEMANY L A， et al. Legal NERC with ontologies， Wikipedia and curriculum learning［C］// Proceedings of the 15th European Chapter of the Association for Computational Linguistics， Volume 2 （Short Papers）. Stroudsburg， PA： Association for Computational Linguistics， 2017： 254-259. 10.18653/v1/e17-2041
8	马建刚，张鹏，马应龙. 基于知识块摘要和词转移距离的高效司法文档分类［J］. 计算机应用， 2019， 39（5）：1293-1298. 10.11772/j.issn.1001-9081.2018102085
	MA J G， ZHANG P， MA Y L. Efficient judicial document classification based on knowledge block summarization and word mover’s distance［J］. Journal of Computer Applications， 2019， 39（5）：1293-1298. 10.11772/j.issn.1001-9081.2018102085
9	马建刚，马应龙. 语义驱动的司法文档学习分类方法［J］. 计算机应用， 2019， 39（6）：1696-1700. 10.11772/j.issn.1001-9081.2018109193
	MA J G， MA Y L. Semantic-driven learning and classification method of judicial documents［J］. Journal of Computer Applications， 2019， 39（6）：1696-1700. 10.11772/j.issn.1001-9081.2018109193
10	GIBAJA E， VENTURA S. Multi‐label learning： a review of the state of the art and ongoing research［J］. WIREs Data Mining and Knowledge Discovery， 2014， 4（6）： 411-444. 10.1002/widm.1139
11	LIU W W， WANG H B， SHEN X B， et al. The emerging trends of multi-label learning［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2021（Early Access）： 1-1.
12	TSOUMAKAS G， KATAKIS I. Multi-label classification： an overview［J］. International Journal of Data Warehousing and Mining， 2007， 3（3）： 1-13. 10.4018/jdwm.2007070101
13	张洛阳，毛嘉莉，刘斌，等. 基于贝叶斯模型的多标签分类算法［J］. 计算机应用， 2016， 36（1）： 52-56， 71. 10.11772/j.issn.1001-9081.2016.01.0052
	ZHANG L Y， MAO J L， LIU B， et al. Multi-label classification algorithm based on Bayesian model［J］. Journal of Computer Applications， 2016， 36（1）： 52-56， 71. 10.11772/j.issn.1001-9081.2016.01.0052
14	ZHU X J. Semi-supervised learning literature survey： TR1530［R］. Madison， WI： University of Wisconsin-Madison， Department of Computer Sciences， 2005： 10.
15	DEVLIN J， CHANG M W， LEE K， et al. BERT： pre-training of deep bidirectional transformers for language understanding［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）. Stroudsburg， PA： Association for Computational Linguistics， 2019： 4171-4186. 10.18653/v1/n19-1423
16	VASWANI A， SHAZEER N， PARMAR N， et al. Attention is all you need［C］// Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook， NY： Curran Associates Inc.， 2017： 6000-6010. 10.1016/s0262-4079(17)32358-8
17	van ROOYEN B， MENON A K， WILLIAMSON R C. Learning with symmetric label noise： the importance of being unhinged［C］// Proceedings of the 28th International Conference on Neural Information Processing Systems. Cambridge： MIT Press， 2015： 10-18.
18	JIANG L， ZHOU Z Y， LEUNG T， et al. MentorNet： learning data-driven curriculum for very deep neural networks on corrupted labels［C］// Proceedings of the 35th International Conference on Machine Learning. New York： JMLR.org， 2018： 2304-2313.
19	NORTHCUTT C， JIANG L， CHUANG I. Confident learning： estimating uncertainty in dataset labels［J］. Journal of Artificial Intelligence Research， 2021， 70： 1373-1411. 10.1613/jair.1.12125

[1]	Yongru QIU, Guangle YAO, Jie FENG, Haoyu CUI. Single image de-raining algorithm based on semi-supervised learning [J]. Journal of Computer Applications, 2022, 42(5): 1577-1582.
[2]	Yongchun BAO, Jianchen ZHANG, Shouxin DU, Junjun ZHANG. Multi-label classification algorithm based on non-negative matrix factorization and sparse representation [J]. Journal of Computer Applications, 2022, 42(5): 1375-1382.
[3]	Wei REN, Hexiang BAI. Multi-label image classification method based on global and local label relationship [J]. Journal of Computer Applications, 2022, 42(5): 1383-1390.
[4]	Yuchang YIN, Hongyuan WANG, Li CHEN, Zundeng FENG, Yu XIAO. One-shot video-based person re-identification with multi-loss learning and joint metric [J]. Journal of Computer Applications, 2022, 42(3): 764-769.
[5]	Jie WU, Shitian ZHANG, Haibin XIE, Guang YANG. Semi-supervised knee abnormality classification based on multi-imaging center MRI data [J]. Journal of Computer Applications, 2022, 42(1): 316-324.
[6]	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.
[7]	ZHANG Shipeng, LI Yongzhong, DU Xiangtong. Intrusion detection model based on semi-supervised learning and three-way decision [J]. Journal of Computer Applications, 2021, 41(9): 2602-2608.
[8]	MAO Mingze, CAO Ruihao, YAN Chungang. Semi-supervised classification algorithm based on weight diversity [J]. Journal of Computer Applications, 2021, 41(9): 2473-2480.
[9]	CAO Yuhong, XU Hai, LIU Sun'ao, WANG Zixiao, LI Hongliang. Review of deep learning-based medical image segmentation [J]. Journal of Computer Applications, 2021, 41(8): 2273-2287.
[10]	Junhua YAN, Ping HOU, Yin ZHANG, Xiangyang LYU, Yue MA, Gaofei WANG. Multiply distortion type judgement method based on multi-scale and multi-classifier convolutional neural network [J]. Journal of Computer Applications, 2021, 41(11): 3178-3184.
[11]	ZHU Yuna, ZHANG Yutao, YAN Shaoge, FAN Yudan, CHEN Hantuo. Protocol identification approach based on semi-supervised subspace clustering [J]. Journal of Computer Applications, 2021, 41(10): 2900-2904.
[12]	LYU Yali, MIAO Junzhong, HU Weixin. Semi-supervised learning algorithm of graph based on label metric learning [J]. Journal of Computer Applications, 2020, 40(12): 3430-3436.
[13]	CHENG Kai, WANG Yan, LIU Jianfei. Semi-supervised learning method for automatic nuclei segmentation using generative adversarial network [J]. Journal of Computer Applications, 2020, 40(10): 2917-2922.
[14]	CHEN Kejia, YANG Zeyu, LIU Zheng, LU Hao. Graph convolutional network model using neighborhood selection strategy [J]. Journal of Computer Applications, 2019, 39(12): 3415-3419.
[15]	REN Fulong, CAO Peng, WAN Chao, ZHAO Dazhe. Grading of diabetic retinopathy based on cost-sensitive semi-supervised ensemble learning [J]. Journal of Computer Applications, 2018, 38(7): 2124-2129.

Recognition of sentencing circumstances in adjudication documents based on abductive learning

基于反绎学习的裁判文书量刑情节识别

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 8

References 19

Related Articles 15

Recommended Articles

Metrics