LeDA: A System for Legal Data Annotation

This paper presents LeDA, a system for Legal Data Annotation. The system offers the functionality of annotating and categorising text spans representing legal concepts that capture the topic of a document, and also supports a meta-annotator to adjudicate the ground truth created by different annotators. Notably, our system supports a dynamic update of the ontology by enabling the creation of new legal concepts. Currently employed to annotate key legal concepts, LeDA aims to construct concept-based semantic representations for tasks such as similar case retrieval, and judgment prediction.