Legal IR and NLP: The History, Challenges, and State-of-the-Art

Abstract

Artificial Intelligence (AI), Machine Learning (ML), Information Retrieval (IR) and Natural Language Processing (NLP) are transforming the way legal professionals and law firms approach their work. The significant potential for the application of AI to Law, for instance, by creating computational solutions for legal tasks, has intrigued researchers for decades. This appeal has only been amplified with the advent of Deep Learning (DL). It is worth noting that working with legal text is far more challenging as compared to the other subdomains of IR/NLP, mainly due to the typical characteristics of legal text, such as considerably longer documents, complex language and lack of large-scale annotated datasets. In this tutorial, we introduce the audience to these characteristics of legal text, and with it, the challenges associated with processing the legal documents. We touch upon the history of AI and Law research, and how it has evolved over the years from relatively simpler approaches to more complex ones, such as those involving DL. We organize the tutorial as follows. First, we provide a brief introduction to state-of-the-art research in the general domain of IR and NLP. We then discuss in more detail IR/NLP tasks specific to the legal domain. We outline the methodologies (both from an academic and industry perspective), and the available tools and datasets to evaluate the methodologies. This is then followed by a hands-on coding/demo session.