A Method for Analyzing News Themes Involving Cases with Integrated Crime Classification

doi:10.19678/j.issn.1000-3428.0068823

Computer Engineering ›› 2025, Vol. 51 ›› Issue (4): 208-216. doi: 10.19678/j.issn.1000-3428.0068823

• Cyberspace Security • Previous Articles Next Articles

A Method for Analyzing News Themes Involving Cases with Integrated Crime Classification

YIN Zhaoliang¹^,²^,³, HUANG Yuxin¹^,², YU Zhengtao¹^,²^,*(), WANG Guanwen¹^,², AI Chuanxian¹^,²^,³

1. Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, Yunnan, China
2. Key Laboratory of Artificial Intelligence in Yunnan Province, Kunming University of Science and Technology, Kunming 650500, Yunnan, China
3. Yunnan Branch of the National Computer Network Emergency Response Technical Team/Coordination Center of China, Kunming 650500, Yunnan, China

Received:2023-11-13 Online:2025-04-15 Published:2024-05-22
Contact: YU Zhengtao

融合罪名分类的涉案新闻主题分析方法

尹兆良¹^,²^,³, 黄于欣¹^,², 余正涛¹^,²^,*(), 王冠文¹^,², 艾传鲜¹^,²^,³

1. 昆明理工大学信息工程与自动化学院, 云南昆明 650500
2. 昆明理工大学云南省人工智能重点实验室, 云南昆明 650500
3. 国家计算机网络应急技术处理协调中心云南分中心, 云南昆明 650500

通讯作者: 余正涛
基金资助:
国家自然科学基金(U21B2027); 国家自然科学基金(61972186); 国家自然科学基金(62266027); 国家自然科学基金(62266028); 云南省重大科技专项(202302AD080003); 云南省重大科技专项(202202AD080003); 202202AD080003(202301AT070393); 202202AD080003(202301AT070471)

Abstract

Abstract:

This paper discusses the significance of crime news topic analysis and identifies the limitations of existing methods. The paper presents a novel topic analysis model called the Bidirectional Encoder Representations from Transformers-based Embedded Crime Topic Model (BERT-ECTM) to address identified issues. The model leverages crime charges from legal documents as supervision signals and combines them with crime news text as input to enhance the accuracy and crime preferences of crime news topic information. The model adopts a BERT-based embedded topic analysis approach to capture contextual semantic features effectively. This paper also introduces a variation inference method that approximates the posterior distribution for improved distribution results, to address the challenge of complex marginal distribution estimation during model training. The proposed model is significantly more effective and accurate than traditional methods in analyzing specific crime news topics.

Key words: text topic extraction, crime classification, BERT-ECTM model, involved preferences, text semantics, semantic feature encoding, variegation inference

摘要：

介绍涉案新闻主题分析的应用场景以及现有方法的不足之处。针对这些不足, 提出一种融合罪名分类的主题分析模型BERT-ECTM。该模型利用法律文书中的罪名信息作为监督信号, 与涉案新闻文本相融合作为主题分析模型的输入, 以提高涉案新闻主题信息的准确性和涉案偏好。为了捕捉上下文语义特征, 采用基于BERT编码的嵌入式主题分析方法, 以提高主题分析的准确性和效果。此外, 在模型训练时, 针对边际分布求解难度较大、复杂程度高这一问题, 结合变分推断的方法, 用后验分布的近似分布来拟合其分布结果。实验结果表明, 在特定的涉案新闻主题分析任务中, 该模型的有效性和准确性相比于现有方法均有明显提升。

关键词: 文本主题抽取, 罪名分类, BERT-ECTM模型, 涉案偏好, 文本语义, 语义特征编码, 变分推断

YIN Zhaoliang, HUANG Yuxin, YU Zhengtao, WANG Guanwen, AI Chuanxian. A Method for Analyzing News Themes Involving Cases with Integrated Crime Classification[J]. Computer Engineering, 2025, 51(4): 208-216.

尹兆良, 黄于欣, 余正涛, 王冠文, 艾传鲜. 融合罪名分类的涉案新闻主题分析方法[J]. 计算机工程, 2025, 51(4): 208-216.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0068823

https://www.ecice06.com/EN/Y2025/V51/I4/208

Figures/Tables 12

Fig.1 Input representation of WoBERT

Fig.2 Structure of the BERT-ECTM thematic analysis model

Fig.3 Network structure of BERT-ECTM

Fig.4 The relationship between the number of topics, perplexity, and coherence

Fig.5 Comparison between product and original product

Fig.6 Word vector distribution of BERT-ETM and BERT-ECTM

References 28

1	BLEI D M , NG A Y , JORDAN M I . Latent Dirichlet allocation. Journal of Machine Learning Research, 2003, 3, 993- 1022.
2	MOODY C E. Mixing Dirichlet topic models and word embeddings to make LDA2Vec[EB/OL]. [2023-10-05]. https://arxiv.org/abs/1605.02019.
3	DIENG A B , RUIZ F J R , BLEI D M . Topic modeling in embedding spaces. Transactions of the Association for Computational Linguistics, 2020, 8, 439- 453. doi: 10.1162/tacl_a_00325
4	DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[EB/OL]. [2023-10-05]. https://arxiv.org/abs/1810.04805.
5	BRAHMA S. Improved language modeling by decoding the past[EB/OL]. [2023-10-05]. https://arxiv.org/abs/1808.05908.
6	王宇晗, 林民, 李艳玲, 等. 基于BERT的嵌入式文本主题模型研究. 计算机工程与应用, 2023, 59 (1): 169- 179.
	WANG Y H , LIN M , LI Y L , et al. Research on embedded text topic model based on BERT. Computer Engineering and Applications, 2023, 59 (1): 169- 179.
7	覃俊, 刘璐, 刘晶, 等. 基于BERT与主题模型联合增强的长文档检索模型. 中南民族大学学报(自然科学版), 2023, 42 (4): 469- 476.
	QIN J , LIU L , LIU J , et al. Long document retrieval model based on the joint enhancement of BERT and topic model. Journal of South-Central Minzu University (Natural Science Edition), 2023, 42 (4): 469- 476.
8	张明芳, 余正涛, 郭军军, 等. 联合罪名预测的涉案新闻重叠实体关系抽取. 南京理工大学学报, 2021, 45 (1): 46- 55.
	ZHANG M F , YU Z T , GUO J J , et al. Overlapping entity relationship extraction of case-involved news through jointing crime prediction. Journal of Nanjing University of Science and Technology, 2021, 45 (1): 46- 55.
9	毛存礼, 梁昊远, 余正涛, 等. 基于神经自回归分布估计的涉案新闻主题模型构建方法. 中文信息学报, 2021, 35 (2): 89- 98.
	MAO C L , LIANG H Y , YU Z T , et al. Topic model of judicial news based on neural autoregressive distribution estimator. Journal of Chinese Information Processing, 2021, 35 (2): 89- 98.
10	韩鹏宇, 高盛祥, 余正涛, 等. 基于案件要素指导的涉案舆情新闻文本摘要方法. 中文信息学报, 2020, 34 (5): 56-63, 73. doi: 10.3969/j.issn.1003-0077.2020.05.009
	HAN P Y , GAO S X , YU Z T , et al. Case-involved public opinion news summarization with case elements guidance. Journal of Chinese Information Processing, 2020, 34 (5): 56-63, 73. doi: 10.3969/j.issn.1003-0077.2020.05.009
11	LIU Y, LAPATA M. Text summarization with pretrained encoders[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing. [S. l.]: ACL, 2019: 3730-3740.
12	NEWMAN D, LAU J H, GRIESER K, et al. Automatic evaluation of topic coherence[EB/OL]. [2023-10-05]. https://aclanthology.org/N10-1012.pdf.
13	STEVENS K, KEGELMEYER P, ANDRZEJEWSKI D, et al. Exploring topic coherence over many models and many topics[C]//Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning. New York, USA: ACM Press, 2012: 952-961.
14	RÖDER M, BOTH A, HINNEBURG A, et al. Exploring the space of topic coherence measures[C]//Proceedings of the 8th ACM International Conference on Web Search and Data Mining. New York, USA: ACM Press, 2015: 399-408.
15	MIMNO D, WALLACH H M, TALLEY E, et al. Optimizing semantic coherence in topic models[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing. New York, USA: ACM Press, 2011: 262-272.
16	SCHNEIDER K M. Weighted average pointwise mutual information for feature selection in text categorization[EB/OL]. [2023-10-05]. https://link.springer.com/content/pdf/10.1007/11564126_27.pdf.
17	YANG Z L, DAI Z H, YANG Y M, et al. XLNet: generalized autoregressive pretraining for language understanding[EB/OL]. [2023-10-05]. https://arxiv.org/abs/1906.08237.
18	HEIDARI M, JONES J H. Using BERT to extract topic-independent sentiment features for social media bot detection[C]// Proceedings of the 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference. Washington D.C., USA: IEEE Press, 2020: 542-547.
19	雷明珠, 邵新慧. 短文本分类模型的优化及应用. 计算机应用研究, 2021, 38 (6): 1775- 1779.
	LEI M Z , SHAO X H . Optimization and application of short text classification model. Application Research of Computers, 2021, 38 (6): 1775- 1779.
20	HAO Y R, DONG L, WEI F R, et al. Visualizing and understanding the effectiveness of BERT[EB/OL]. [2023-10-05]. https://arxiv.org/abs/1908.05620v1.
21	SALTON G , WONG A , YANG C S . A vector space model for automatic indexing. Communications of the ACM, 1975, 18 (11): 613- 620. doi: 10.1145/361219.361220
22	GUO C H , LU M L , WEI W . An improved LDA topic modeling method based on partition for medium and long texts. Annals of Data Science, 2021, 8 (2): 331- 344. doi: 10.1007/s40745-019-00218-3
23	TAN B W, YANG Z C, AI-SHEDIVAT M, et al. Progressive generation of long text with pretrained language models[EB/OL]. [2023-10-05]. https://arxiv.org/abs/2006.15720v2.
24	ZHAO L, XU J J, LIN J Y, et al. Graph-based multi-hop reasoning for long text generation[EB/OL]. [2023-10-05]. https://arxiv.org/abs/2009.13282v1.
25	CHEN K, FU G, CHEN Q, et al. A large-scale Chinese long-text extractive summarization corpus[EB/OL]. [2023-10-05]. https://ieeexplore.ieee.org/document/9414946.
26	CHANGPINYO S, SHARMA P, DING N, et al. Conceptual 12M: pushing web-scale image-text pre-training to recognize long-tail visual concepts[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2021: 3558-3568.
27	LI W L , JIN B , QUAN Y . Review of research on text sentiment analysis based on deep learning. OALib, 2020, 7 (3): 1- 8.
28	薛琪, 孟祥福, 张峰, 等. HLMGAN: 分层学习的多奖励文本生成对抗网络. 云南大学学报(自然科学版), 2022, 44 (1): 64- 72.

Please choose a citation manager

Content to export