Al & Law: Application on Argumentation4 min read

Arguments, those are found in judicial decisions and case files can often be rather complex so that understanding the link of relationships becomes difficult. There is a strong potential for computers to provide a means of addressing this problem, and the development of argument structuring systems might help (Prakken, 2008). Currently, Al (Artificial Intelligence) & Law has addressed this need in research on argument structuring systems, which are a method of self-styled sense-making systems. Those type of systems circumvents the knowledge acquisition bottleneck because they do not have the knowledge which they apply to solve a problem. Instead, they support humans in making sense of a problem, by providing tools for structuring; typically visualizing the problem and the user’s reasoning in solving it. Some sense-making systems also help to communicate with different people working on the same problem.

Al and Law researchers build computational models of legal reasoning, which are computer programs that perform or simulate legal reasoning (Ashley, 2013). “If the inputs to the program are a description of a legal problem, its outputs might include a solution to the legal problem and an explanation”. Legal problem-solving repeatedly requires reasoned explanations as well as a decision or prediction. Undeniably, the argument supporting a decision may be more important than the decision; otherwise, a human user cannot judge whether to rely on the advice. The focus on explanations and arguments is a key area where Al and Law contribute to Al research.

Likewise, a computational model can be described just like computer programs, in terms of the inputs to the program, its outputs, and the intervening steps that transform the former into the latter. The intervening steps or algorithm transforms the inputted problem description into the outputs.

In specifying an Al program’s input/output (1/O) behavior, three questions are especially relevant:

  1. Knowledge representation and search: How are the inputs, and the information used to analyze them, represented in a problem space that can be searched systematically?
  2. Inference control mechanism: What governs the search for solutions in terms of efficiency and relevance?
  3. Learning: In order to improve its performance, how can the program learn from its errors and successes and from other sources of information? In the case of Al and Law programs, only a few accept problems inputted as natural language texts or reason with cases or legal rules expressed in text (Garfield, 2013).

The emphasis will be on the ways in which the application/program search analytically for solutions and can learn, for example, how to detect factors in the texts of cases or to classify the texts of statutory provisions. In principle, and often in practice, an Al or Al and Law program’s problem-solving behavior can be assessed in terms of relevance and performance measures similar to those applied to human problem-solving for instance, predictive accuracy, recall, precision, coverage, as well as the explanatory adequacy of the outputs given the inputs.

Consequently, the structure or model would capture the main issues, positions and arguments taken by the parties with respect to the issues, the available evidence related to them, and so on. Incoming documents could be indexed according to this structure and new documents either outgoing documents or internal analyses of a case could be drafted consistent with the same structure and linked to relevant background documents, for example, case law, statutes, journal articles, testimonies etc. Similarly, work on argumentation schemes can further augment the usefulness of such systems. When constructing arguments, argumentation schemes provide a repertoire of forms of argument to be considered, and a template prompting for the pieces that are needed; when attacking arguments, they provide a set of critical questions that can identify potential weaknesses in the opponent’s case.

In conclusion, argument structuring systems have uses in areas where the clear presentation of the argument is of principal importance, for example, preliminary fact investigation and case management. In all these instances, the usefulness of such systems might be increased by integrating them with documentary sources. For instance, when used for case management, the software could allow the user to structure a collection of case-related documents in terms of the argumentation structure of a case. Since argument structuring systems avoid the knowledge acquisition bottleneck, they scale up to realistic size more easily than knowledge based argumentation systems.

 

References:

Henry Prakken, ‘AI & Law on Legal Argument: Research Trends and Application Prospects’ (2008) 5, 3.

Kevin D. Ashley, ‘Teaching Law and Digital Age Legal Practice with an AI and Law Seminar’ (2013) 88 Chi.-Kent L. Rev. 783.

Sadia Sharmin

Founder CloudOnex, Studied LL.M in Intellectual Property Law at Uppsala University