Sentiment Analysis of Canadian Maritime Case Law: A Sentiment Case Law and Deep Learning Approach

This study employs machine and deep learning methods to expedite legal proceedings to retrieve legal data, classify, review, and predict judgments. Our study adds significantly to the literature because many countries' judicial systems have backlogs that cause delays in justice. The sentiment analysis framework employs deep, distributed, and machine learning to provide access to statutes, laws, and cases, allowing Canadian maritime judges to resolve cases more efficiently. The proposed LSTM+CNN model demonstrated promising results in extracting sentiments and records from various devices and providing practical guidance. As a result, the model can be applied to other systems that adhere to the common-law framework.本研究采用机器学习与深度学习技术，旨在加速法律程序，实现法律数据的获取、分类、审查及裁判结果预判。鉴于诸多国家的司法系统均存在引发司法延误的案件积压现象，本研究对相关学术研究具有重要的补充意义。本研究提出的情感分析框架融合深度学习、分布式学习与机器学习技术，可实现成文法、法律法规与司法判例的检索，助力加拿大海事法官更高效地审结案件。本次提出的长短期记忆网络与卷积神经网络（LSTM+CNN）混合模型，在从各类设备中提取情感信息与相关记录、提供实用指导方面展现出优异效果。据此，该模型可推广应用于其他采用普通法框架的司法系统。