基于预训练语言模型的案件要素识别方法(A Method for Case Factor Recognition Based on Pre-trained Language Models)

no code implementations • CCL 2020 • Haishun Liu, Lei Wang, Yanguang Chen, Shuchen Zhang, Yuanyuan Sun, Hongfei Lin

案件要素识别指将案件描述中重要事实描述自动抽取出来, 并根据领域专家设计的要素体系进行分类, 是智慧司法领域的重要研究内容。基于传统神经网络的文本编码难以提取深层次特征, 基于阈值的多标签分类难以捕获标签间依赖关系, 因此本文提出了基于预训练语言模型的多标签文本分类模型。该模型采用以Layer-attentive策略进行特征融合的语言模型作为编码器, 使用基于LSTM的序列生成模型作为解码器。在“CAIL2019”数据集上进行实验, 该方法比基于循环神经网络的算法在F1值上最高可提升7. 6%, 在相同超参数设置下比基础语言模型(BERT)提升约3. 2%。

Pre-trained Mixed Integer Optimization through Multi-variable Cardinality Branching

no code implementations • 21 May 2023 • Yanguang Chen, Wenzhi Gao, Dongdong Ge, Yinyu Ye

We propose a new method to accelerate online Mixed Integer Optimization with Pre-trained machine learning models (PreMIO).

Learning Theory

Joint Entity and Relation Extraction for Legal Documents with Legal Feature Enhancement

no code implementations • COLING 2020 • Yanguang Chen, Yuanyuan Sun, Zhihao Yang, Hongfei Lin

We propose a Legal Triplet Extraction System for drug-related criminal judgment documents.

Joint Entity and Relation Extraction Multi-Task Learning +4

New Approaches for Calculating Moran's Index of Spatial Autocorrelation

2 code implementations • 12 Jun 2016 • Yanguang Chen

Spatial autocorrelation plays an important role in geographical analysis, however, there is still room for improvement of this method.

Physics and Society Methodology