Leveraging Label Semantics and Correlations for Judgment Prediction

Automatic judgment prediction is a classic problem in legal intelligence, which aims to predict the relevant violated articles based on the fact descriptions. Generally, both semantics and relations of articles are valuable information to solve this problem. However, previous work usually threats this problem as a classification task while these two types of information are not well explored, which makes previously proposed methods less effective. In this paper, we design a novel Graph-Based Label Matching Network (GLAM for short) to address this issue. Specifically, GLAM first builds a heterogeneous graph to capture both semantics and correlations among articles. Based on this, a graph convolutional network is then utilized to learn robust article representations. Finally, a matching model is applied between article representations and fact representations to generate the matching score for judgment prediction. Experimental results on two real-world judicial datasets demonstrate that our model has more significant effect on judgement prediction than the state-of-the-art methods.

[1]  Joan Bruna,et al.  Spectral Networks and Locally Connected Networks on Graphs , 2013, ICLR.

[2]  Philip S. Yu,et al.  A Comprehensive Survey on Graph Neural Networks , 2019, IEEE Transactions on Neural Networks and Learning Systems.

[3]  Jeffrey Pennington,et al.  GloVe: Global Vectors for Word Representation , 2014, EMNLP.

[4]  Xueqi Cheng,et al.  A Deep Architecture for Semantic Matching with Multiple Positional Sentence Representations , 2015, AAAI.

[5]  Zhiyuan Liu,et al.  Legal Judgment Prediction via Topological Learning , 2018, EMNLP.

[6]  Larry P. Heck,et al.  Learning deep structured semantic models for web search using clickthrough data , 2013, CIKM.

[7]  Yuan Luo,et al.  Graph Convolutional Networks for Text Classification , 2018, AAAI.

[8]  W. Bruce Croft,et al.  Semantic Matching by Non-Linear Word Transportation for Information Retrieval , 2016, CIKM.

[9]  Jimmy J. Lin.  An exploration of the principles underlying redundancy-based factoid question answering , 2007, TOIS.

[10]  Zhiyuan Liu,et al.  Few-Shot Charge Prediction with Discriminative Legal Attributes , 2018, COLING.

[11]  Geoff Holmes,et al.  Classifier chains for multi-label classification , 2009, Machine Learning.

[12]  Pengfei Wang,et al.  Modeling Dynamic Pairwise Attention for Crime Classification over Legal Articles , 2018, SIGIR.

[13]  Benjamin E. Lauderdale,et al.  The Supreme Court's Many Median Justices , 2012, American Political Science Review.

[14]  Max Welling,et al.  Semi-Supervised Classification with Graph Convolutional Networks , 2016, ICLR.

[15]  Joan Bruna,et al.  Deep Convolutional Networks on Graph-Structured Data , 2015, ArXiv.

[16]  Zhi-Hong Deng,et al.  Inter-Weighted Alignment Network for Sentence Pair Modeling , 2017, EMNLP.

[17]  Yoon Kim,et al.  Convolutional Neural Networks for Sentence Classification , 2014, EMNLP.

[18]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[19]  Zhiyuan Liu,et al.  Automatic Judgment Prediction via Legal Reading Comprehension , 2018, CCL.

[20]  Hang Li,et al.  Convolutional Neural Network Architectures for Matching Natural Language Sentences , 2014, NIPS.

[21]  Xavier Bresson,et al.  Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering , 2016, NIPS.

[22]  Nikolaos Aletras,et al.  Predicting judicial decisions of the European Court of Human Rights: a Natural Language Processing perspective , 2016, PeerJ Comput. Sci..

[23]  Dongyan Zhao,et al.  Learning to Predict Charges for Criminal Cases with Legal Basis , 2017, EMNLP.

[24]  Kenneth Ward Church,et al.  Word Association Norms, Mutual Information, and Lexicography , 1989, ACL.

[25]  Jianxin Li,et al.  Large-Scale Hierarchical Text Classification with Recursively Regularized Deep Graph-CNN , 2018, WWW.

[26]  Zhi-Hua Zhou,et al.  Multilabel Neural Networks with Applications to Functional Genomics and Text Categorization , 2006, IEEE Transactions on Knowledge and Data Engineering.