摘要
针对当前数据驱动的方法罪名预测准确率较低的难题,提出新的罪名预测方法。文章利用生成对抗网络(GAN)对案件事实文本进行数据增强,扩充了训练样本。并对BERT模型进行改进,引用卷积神经网络和双向长短期记忆网络提取案件的局部特征以及前后文特征,提高重点词语在案件事实中的作用。实验结果表明,该方法相比于其他罪名预测模型,具有更高的准确率,准确率、召回率、F1值分别为0.9416,0.9412,0.9403。
Aiming at the problem that the current data-driven methods have low accuracy in predicting legal charges,a new online crime prediction method is proposed.In this paper,we use Generative Adversarial Networks(GAN)to perform data augmentation on case fact texts to expand the training samples.On the basis of the original BERT model,CNN and Bi-LSTM are used to extract the local features and contextual features of the case facts,which improves the role of key words in the case facts,thereby improving the accuracy of crime prediction.The experimental results show that this method has higher accuracy than other crime prediction models,and the accuracy,recall,and F1 score are 0.9416,0.9412,and 0.9403,respectively.
作者
白昌前
代晓
张岸
BAI Chang-qian;DAI Xiao;ZHANG An(School of Cyberspace Security and Information Law,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;College of Public Administration,Chongqing University,Chongqing 400044,China)
出处
《电脑与信息技术》
2023年第1期37-40,共4页
Computer and Information Technology
关键词
法律罪名预测
BERT
生成对抗网络
长短期记忆网络
crime prediction
BERT
generative adversarial networks
long short-term memory network
