Using WikiData for Handling Legal Rule Exceptions: Proof of Concept

Since social expectations and circumstances always change, it is challenging to refine formalized legal rules for handling new exceptions driven from cases. Recent research has explored using linked open datasets for handling exceptions. However, most research uses domain-dependent datasets developed by specific authorities. To handle legal rule exceptions, which typically need knowledge from various domains and need to be updated in response to social changes, we present in this paper a framework for handling legal rule exceptions using Wikidata. Wikidata is a linked open dataset retrieved from various Wikimedia projects with collaborative edits. The framework aims to guide the user to find the boundary of exceptional situation using descendent subclass hierarchies in Wikidata. As Wikidata cannot always handle the correspondence between legal concepts describing the antecedent of the rule and factors describing the case, the framework also helps guiding the user in revising rules or knowledge bases to restore the correspondence. We demonstrate the framework using cases from the European Court of Human Rights, and discuss possible improvements of this framework by integrating it with large language models such as ChatGPT.