You are here

Textual information extraction and ontologies for legal case-based reasoning

This talk gave a brief overview of current developments and prospects in two related areas of the legal semantic web for legal cases - textual information extraction and ontologies. Textual information extraction is a process of automatically annotating and extracting textual information from the legal case base (precedents), thereby identifying elements such as participants, the roles the participants play, the factors which were considered in arriving at a decision, and so on. The information is valuable not only for search (to find applicable precedents), but also to populate an ontology for legal case-based reasoning. An ontology is a formal representation of key aspects of the knowledge of legal professionals with which we can reason (e.g. given an assertion that something is a legal case, we can infer other properties) and with respect to which we can write rules (e.g. reasoning using case factors to arrive at a legal decision). Since it is expensive to manually populate an ontology (meaning to read cases and input the data into the ontology), we use textual information extraction to automatically populate the ontology. The talk was concluded with an appeal for open source, collaborative development of legal knowledge systems among partners in academia, industry, and government.

Presentation Type: 
Presentation Audio: 
Audio Size: 
Presentation Visual: