Dr. Elisabetta Fersini of the University of Milano-Bicocca’s Laboratory for Models in Decision Making and Data Analysis (MIND) has posted The JUMAS Experience: Extracting Knowledge From Judicial Multimedia Digital Libraries, on the VoxPopuLII Blog, published by the Legal Information Institute at Cornell University Law School.
In this post, Dr. Fersini describes JUMAS (JUdicial MAnagement by digital libraries Semantics), an innovative platform, funded by the European Commission, for improving access to the content of digital audio and video recordings of court proceedings. Now being tested in courts in Italy and Poland, JUMAS is designed for use by judges and other court personnel. Dr. Fersini’s post describes each component of JUMAS:
- An automatic transcription system;
- Software for recognizing emotional states of depicted persons;
- Software for recognizing and annotating human behavior depicted in videos;
- Deception detection software;
- An information extraction system;
- An information retrieval system, featuring query expansion;
- An ontology of concepts pertaining to court proceedings;
- Software that enables users to add semantic annotations to audio and video;
- A user interface that enables “the user to perform queries on contents, and jump directly to relevant parts of media files”; and
- Software that automatically summarizes audio and video content.
This post should be of interest to those developing, managing, or administering court technology; the litigation technology community; scholars and consultants who work with video and audio of court proceedings; and all those interested in improving digital access to proceedings of legal tribunals.