COMPUTATIONAL JURISPRUDENCE
| Author | ZHANG Ni1 LLB (Sichuan University), LLM (Sichuan University), JD (Southwestern University of Finance and Economics); Associate Research Fellow, Library of Sichuan University. |
| Publication year | 2021 |
| Citation | (2021) 33 SAcLJ 10355 |
| Published date | 01 December 2021 |
| Date | 01 December 2021 |
1 In a digital society, our movements can be traced and compiled to produce a comprehensive picture of both individual and group behaviour. Driven by legal data, we are both the executors and the rulemaking participants of legal regulations in social machines,2 and our behaviour can be computed and predicted using algorithmic power.3 As an interdisciplinary field involving social and computer sciences,
2 Computational social science has made considerable progress in such fields as economics, sociology and politics over the past ten years, and also has great influence on other social subjects. Legal systems have evolved into complex systems with diverse institutions (eg, legislatures, agencies and courts), norms (eg, due process, equality and fairness), actors (eg, legislators, bureaucrats and judges) and instruments (eg, regulations, injunctions and taxes) that are interconnected through stochastic processes (eg, trials, negotiations and rulemaking) with feedback mechanisms (eg, appeals to higher courts and judicial review of legislation).6 As the legal process becomes increasingly complicated, data-driven legal systems can provide an important mechanism to better manage conflicts in society. In the era of big data and artificial intelligence (“AI”), the use of paperless lawsuits and audio-visual recordings in courts in addition to judicial reports can be used to track the process of litigation, and to compile comprehensive portraits of litigants and the court. It may even be possible to use this information to predict judicial sentences.
3 While quantitative methods have been applied to the law for more than 70 years now, in many instances, these applications have not been considered as an independent research subject. However, with the development of AI and machine learning techniques, the investigation of legal relationships and interactions through quantitative methods — which is described here as “computational jurisprudence” — is expected to become an important and independent discipline in itself. “Computational jurisprudence” will be further defined, and its aims, values and objectives set out, in Part II of this article. Why computational jurisprudence will be an emerging field of interdisciplinary study will also be discussed in Part III. And in Part IV, it will be explained how to build a computational model and the dilemmas and problems faced in doing so. The article will wrap up with an explanation as to how the use of the latest computational techniques will greatly change the role of law in society.
4 The application of computer science in fields such as civil law, criminal law and administrative law has given rise to a promising area of interdisciplinary research in legal science and is referred to as “computational jurisprudence” that is laying more stress on jurisprudence than computational law. What is computational jurisprudence?
5 Wyner first introduced the concept of computational jurisprudence in 2005 and described how it could perform a logical analysis of the digital version of laws and regulations, as well as the rationality of the automatic detection of laws and regulations.7 This is consonant with the view of Michael Genesereth, the Research Director of CodeX — the Stanford Center for Legal Informatics — who believes that computers can be used to understand legal rules and design automatic legal decision systems, such as smart or computed contracts,8 including legal retrieval technology, e-discovery software, contract analysis, contract management systems, lawyer-matching systems, etc. In 2017, a Chinese scholar, Qian Ningfeng, inferred that computational jurisprudence represents the next logical development of the law in the era of big data.9 But these views do not clearly define what computational jurisprudence is.
6 In 2015, Goodenough analysed the techniques used in computational jurisprudence which he called Law 1.0, 2.0 and 3.0, according to the stages of the development of AI technology. In Law 1.0 and Law 2.0, the main focus is the improvement of information processing capabilities within the current legal system using technologies such as computer-aided legal research, process management, and expert systems for semantic processing. In Law 3.0, existing systems will be largely changed or replaced with a new system driven by the power of computational technology, such as communications, modelling and execution.
7 In 2015, the author co-authored a book named Introduction to Computational Jurisprudence10 and sought to formulate a basic definition of computational jurisprudence that is based on technologies such as quantitative analysis, knowledge graphs and AI modelling. The author concluded that computational jurisprudence encompasses many areas such as the scientific research of legislation, the evaluation of the implementation of laws and its impact on economic and social development, and its use in the verification of the theories of law, the rationality of law, the assessment of evidence in criminal investigations, and its deployment in practical legal aid systems.
8 With the further development of latest technologies and its wider use and acceptance in not just the law but also in computational social sciences, it is apt to review the definition of computational jurisprudence to reflect the progress that has been made so far. In addition, breakthroughs made in new techniques, especially in AI, can now support the analysis of online real-time information in the law.
9 Therefore, “computational jurisprudence” is redefined here as the discovery of laws or principles behind legal systems through the use of intelligent systems that process real-time information to investigate legal relationships and interactions through simulations and modelling. Computational jurisprudence focuses on the impact that the use of multiple computational methods, including AI and machine learning, data-mining and other methods, will have on the legal system, as well as on improving our understanding of the realities which underlie the legal system and on which the law must act. The aims of computational jurisprudence are not only concerned with efficiency, but also the completion of legal theory and the discovery of the operative rules or underlying principles behind the statutes and judicial cases. Computational jurisprudence belongs to the family of computational social sciences, such as computational economics, computational linguistics and computational psychology. It is hoped that computational jurisprudence will enrich legal theory, improve judicial efficiency, reduce conflicts within legal systems, and improve the impartiality and effectiveness of legislative and judicial practices. Finally, it is hoped that computational jurisprudence will promote the harmonisation of common law and civil law systems, and reduce the risks brought by application of new technologies.11
10 The core idea of computational jurisprudence therefore is the integration of computational thinking and legal reasoning. Through computational simulation, AI and complex statistics, AI systems are used to analyse and simulate legal issues to obtain a deeper understanding of legal reasoning, legal systems, legal institutions and the role of law in society. Researchers will no longer deal only with experiences and theories, nor do they simply learn from and imitate natural science. Instead, they will analyse behaviours and social structures through the real-time calculation of observations regarding the data subject and discover patterns that would lead to in-depth legal developments. Computational jurisprudence will build a bridge between legal problems and computational techniques, while also breaking through some of the difficulties in cross-referencing from basic theory, experimental methods and domain application.
11 With an increasing number of technologies being applied in legal fields, the relationship between law and science has become closer. The rubric of computational jurisprudence encompasses computational thinking, law, AI and big data processing as an interdisciplinary discipline of law and computer science. How then is computational jurisprudence different from other disciplines and areas such as jurimetrics, AI and the law and legal informatics?
12 Speculation about the potential benefits of applying methods from computational science and AI research to automate parts of the law date back at least to the mid-1940s.12 This heralded the introduction of jurimetrics, which emphasises the application of quantitative methods, especially statistics, to the law.13 In 1958, Lucien first proposed to informatise legal science, or to be specific, to establish automatic retrieval models for legal documents or cases as well as judges' decisions. The question of how to build rule-based and case-based legal reasoning models (expert systems) has always been the focus of research on AI and the law. And since 1997, with the application of AI and big data mining, legal informatics has emerged as a branch of informatics science.14 Legal informatics emphasises the application of informatics techniques for the retrieval of legal information and for the delivery of
13 While these disciplines were initiated at different times, the research content for all these disciplines is related to research that focuses on the law and uses computational methods for its...
To continue reading
Request your trial