Synthetic Legal Precedent Structures: Lévy Flight

dimensions of Lévy flight

2

3

4

5

α parameter for Lévy flight

2.5

number of cases

50

mean number of out citations per case

4

standard deviation of number of out citations per case

0.1

distance function

Min[Abs[#1-#2]]&

In common-law systems, legal cases tend to cite one another. This Demonstration synthesizes a legal precedent structure in which cases are given up to five features as well as a time of decision. Cases cite preceding cases whose features most closely resemble themselves, with "closeness" being defined in a user-selectable manner. The cases in this Demonstration are generated using a Lévy flight model that attempts to reflect the notion that cases tend to be clustered together rather than randomly filling some "judicial space." It visualizes the Lévy flight where possible and shows the legal precedent network.