Deep Learning Techniques for Legal Text Summarization

Abstract

Summarizing Legal Text data is a significant problem due to the extensive length and complexity involved in analyzing it. The advent of deep neural networks and their demanding application in Natural Language Processing(NLP) paves the way to conquer the legal domain text data. This paper proposes a systematic comparison of various deep learning strategies applied in summarizing Legal Texts. We begin the study with the introduction and evolution of different deep learning models used in text summarization. Through the work, we have identified the importance of various preparation mechanisms within deep learning to enhance the legal text summarization process. The work also focuses on the review of existing deep learning models used recently in the legal domain for different types of summarization. Models employing the sequence to sequence neural network architecture has gained much importance. This architecture progressed with transfer learning to fine-tune the state of the art pre-trained language models, leading to better accuracy in summarization.