Abstract:Legal judgment and prediction is a classic traditional task combining artificial intelligence with law. It aims to train a machine judge model to predict the final charges automatically according to the fact descriptions in criminal cases. In recent years, with the successful usage of deep learning techniques on NLP tasks. The researchers explored the application of LSTM model, neuron model and attention mechanism to predict the charges and achieved good results. However, For those crimes whose facts are similar to each other, the existing algorithms can't get well performance. To address these issues, we propose a novel charge prediction model based on Bi-attention LSTMCNN Hybrid Neural Network (BLHNN) to establish several discriminative properties of charges as internal mapping between fact descriptions and charges. More specifically, first, word2vec algorithm is used to transform the document feature words into fixed dimension word vectors, then TF-IDF value of each feature word in the document is calculated, and TF-IDF value of each feature word is used as the weight to encode the word vectors of all feature words. Then, in order to solve the problem that convolution neural network (CNN) can not get the global features of the text and Bi-LSTM can not focus on the local features of the text in text classification, we adopt a hybrid model including CNN and Bi-LSTM. At the same time, the attention mechanism is introduced to solve the problem of not paying attention to the features of key words in text classification. Finally, the experimental results show that the proposed model can get more accurate text feature information and improve the accuracy of text classification.

Interpretable Charge Prediction for Criminal Cases with Dynamic Rationale Attention

Interpretable Rationale Augmented Charge Prediction System.

Learning to Predict Charges for Criminal Cases with Legal Basis

Creating Auxiliary Representations from Charge Definitions for Criminal Charge Prediction

Interpretable Charge Prediction for Legal Cases Based on Interdependent Legal Information

Interpretable Charge Predictions for Criminal Cases: Learning to Generate Court Views from Fact Descriptions

Charge Prediction with Legal Attention

Few-Shot Charge Prediction with Discriminative Legal Attributes.

What Makes a Charge? Identifying Charge-Discriminative Facts with Legal Elements.

Do Charge Prediction Models Learn Legal Theory?

Knowledge-aware Method for Confusing Charge Prediction

Legal Charge Prediction Via Bilinear Attention Network

Learning Defendant-aware Label Representation for Multi-Defendant Charge Prediction

A hybrid neural network model for predicting charges by incorporating legal knowledge

Learning To Predict Charges For Judgment With Legal Graph

Leveraging Knowledge-Aware Method for Charge Prediction.

BLHNN: A Novel Charge Prediction Model Based on Bi-Attention LSTM-CNN Hybrid Neural Network

A Multi-Source Heterogeneous Knowledge Injected Prompt Learning Method for Legal Charge Prediction

Towards Interactivity and Interpretability: A Rationale-based Legal Judgment Prediction Framework.

ADAN: an Intelligent Approach Based on Attentive Neural Network and Relevant Law Articles for Charge Prediction.

NeurJudge: A Circumstance-aware Neural Framework for Legal Judgment Prediction.