Re20:读论文 What About the Precedent: An Information-Theoretic Analysis of Common Law

简介: Re20:读论文 What About the Precedent: An Information-Theoretic Analysis of Common Law

1. Background


  1. Halsbury’s:认为先例的arguments是重点(实验结果:√)

Goodhart’s:认为先例的事实是重点

(前者比后者更强调推理,后者的假设是如果不想和先例一样,说先例事实不同比推翻先例更容易)


  1. ratio decidendi:判决理由,文中指以先例的arguments或者facts作为判决理由。


2. 模型


image.png

a507d8e607b446c5b6970f2bbb4e3361.png

5ba11c3bba8c4310862c572d7a467934.png

image.png

image.png


用神经网络来近似:

image.png


3. 实验


3.1 数据集

ECHR数据集,案例文本可切分为facts, arguments(法官的解释) and outcome. Arguments cite relevant cases, also known as the precedent

子语料:包含去重的案例引用关系


3.2 实验设置

Longformer

超参:由于资源有限,隐藏层维度50,batch size 16,将每个案例truncate到512个tokens。

4 Nvidia P100 16GiB GPU

最多6小时


3.3 实验结果

image.png

交叉熵小、MI大

image.png

6b1e04dab78e4f3382f6fa7d4bf52dee.png

探讨了一些法条跟别的法条不一样的情况,比如因为发生太晚所以被truncate了……

相关文章
|
算法 计算机视觉 知识图谱
ACL2022:A Simple yet Effective Relation Information Guided Approach for Few-Shot Relation Extraction
少样本关系提取旨在通过在每个关系中使用几个标记的例子进行训练来预测句子中一对实体的关系。最近的一些工作引入了关系信息
121 0
|
人工智能 自然语言处理 算法
UIE: Unified Structure Generation for Universal Information Extraction 论文解读
信息提取受到其不同目标、异构结构和特定需求模式的影响。本文提出了一个统一的文本到结构生成框架,即UIE,该框架可以对不同的IE任务进行统一建模,自适应生成目标结构
488 0
|
机器学习/深度学习 算法 数据挖掘
Re17:读论文 Challenges for Information Extraction from Dialogue in Criminal Law
Re17:读论文 Challenges for Information Extraction from Dialogue in Criminal Law
Re17:读论文 Challenges for Information Extraction from Dialogue in Criminal Law
|
机器学习/深度学习 数据采集 人工智能
Re10:读论文 Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous gr
Re10:读论文 Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous gr
Re10:读论文 Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous gr
|
机器学习/深度学习 自然语言处理 PyTorch
Re6:读论文 LeSICiN: A Heterogeneous Graph-based Approach for Automatic Legal Statute Identification fro
Re6:读论文 LeSICiN: A Heterogeneous Graph-based Approach for Automatic Legal Statute Identification fro
Re6:读论文 LeSICiN: A Heterogeneous Graph-based Approach for Automatic Legal Statute Identification fro
|
搜索推荐 PyTorch 算法框架/工具
Re30:读论文 LegalGNN: Legal Information Enhanced Graph Neural Network for Recommendation
Re30:读论文 LegalGNN: Legal Information Enhanced Graph Neural Network for Recommendation
Re30:读论文 LegalGNN: Legal Information Enhanced Graph Neural Network for Recommendation
|
自然语言处理
Re26:读论文 Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks
Re26:读论文 Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks
Re26:读论文 Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks
|
机器学习/深度学习 算法 数据挖掘
Re18:读论文 GCI Everything Has a Cause: Leveraging Causal Inference in Legal Text Analysis
Re18:读论文 GCI Everything Has a Cause: Leveraging Causal Inference in Legal Text Analysis
Re18:读论文 GCI Everything Has a Cause: Leveraging Causal Inference in Legal Text Analysis
|
机器学习/深度学习 计算机视觉
Re14:读论文 ILLSI Interpretable Low-Resource Legal Decision Making
Re14:读论文 ILLSI Interpretable Low-Resource Legal Decision Making
Re14:读论文 ILLSI Interpretable Low-Resource Legal Decision Making
|
SQL 存储 算法
《Optimization of Common Table Expressions in MPP Database Systems》论文导读
Optimization of Common Table Expressions in MPP Database Systems
《Optimization of Common Table Expressions in MPP Database Systems》论文导读