Predictions with Uncertainty to Support Fair Outcomes in Online Legal Disputes

Alternative dispute resolutions systems are not uncommon in Australian family law, however to date these systems are largely negotiation based and are not designed for producing judicially fair outcomes. This paper proposes an online dispute resolution approach that aims to support divorcees to resolve property issues in a manner that is consistent with orders a judge would make if the matter was heard in court. The approach integrates a protocol for online dispute dialogue with an argument based model of judicial reasoning to structure the dispute. The likelihood of alternates outcomes is predicted with a series of Bayesian belief networks.