Reusing Cases to Automatic Index Assignment from Textual Documents

This paper describes one solution developed to convert textual documents into formlike representations of cases. The experiences described by cases are textual descriptions of legal decisions. Indexing vocabulary and assignment theory contributed in gathering expert knowledge to define attributes and values as well as the required elements to employ template mining. Most index values are automatically extracted by the use of template mining. The multi-purpose index Theme is automatically assigned by reusing cases through an elaboration process. Seed cases are used to indicate values if the new case is a partial match to one in the case base.