September 15th, 2021
1:00 PM - 2:00 PM EDT
Are portable models, the latest hype in eDiscovery, worth considering over human expertise under the continuous active learning (CAL) model?
Portable models—aka pre-trained AI models based on supervised machine learning models trained on one set of eDiscovery data and applied to another—come with risks and rewards in eDiscovery document review.
Based on a peer-reviewed study, OpenText lead data scientist Dr. Jeremy Pickens and head eDiscovery strategy consultant Tom Gricks will explore the questions and present empirical findings on the true value and effectiveness of portable models versus the risks and trade-offs.
Learn the answers to the following questions: Do portable models…
- Help with the “cold start” problem (e.g., increase precision at 80% recall relative to random seeding)?
- Find more seeds, initially?
- Increase precision at 80% recall relative to human effort?
- Show significant and sustained advantages over human expertise?
Register now to get the latest research.
Jeremy Pickens, Ph.D.
Principal Data Scientist
Jeremy is one of the world’s leading information retrieval scientists and a pioneer in the field of collaborative exploratory search. He has seven patents and patents pending in the field of search and information retrieval. Jeremy spearheads the development of Insight Predict. His ongoing research focuses on methods for continuous learning, and the variety of technology assisted review workflows that are only possible with this approach.
Lead Strategy Consultant
Tom is a prominent eDiscovery attorney, leading authority on the use of TAR for litigation and investigations, and eDiscovery Special Master. Tom advises clients on how to optimize machine learning and analytics techniques on investigations, has developed the managed review program for internal, compliance and regulatory investigations, and develops investigation-specific applications and TAR protocols. He brings more than 30 years’ legal and eDiscovery expertise to clients.