Technology-assisted review software, built on a powerful analytics engine, helps you find your most relevant documents faster.
RelativityOne’s active learning workflow continuously learns what’s important to your matter to quickly get to the heart of it.
As you code documents, RelativityOne will keep a pulse on coding decisions in real time, constantly refining its understanding of what’s responsive to get smarter as the review progresses.
Simply code documents and the most relevant ones will be served up next, so reviewers can always take advantage of the system’s most current understanding.
With little administrative support required—and no need to create training sets—you can stand your project up quickly and let the active learning engine do the heavy lifting.
Choose the active learning workflow that’s right for your unique project. The prioritized review queue serves up the most responsive documents first, so you can get the most important ones to your reviewers right away. The coverage queue delivers documents the engine is most unsure about, helping it understand relevancy and allowing you to review fewer documents.
RelativityOne’s sample-based learning workflow uses sampling to slice a document set in various ways, making sure you get coverage across all of your documents as you’re training the system to make decisions.
RelativityOne leverages a seed set of human-coded documents to train the system, giving you a jump start on your review. Based on that, coding decisions are suggested for the remaining document universe, so you can immediately begin QC to refine the system’s understanding.
Easily sort your documents into groups, so you can have responsive documents reviewed by the most qualified experts and uncategorized ones passed to other reviewers for QC.
RelativityOne’s sample-based learning workflow has been proven effective by hundreds of organizations and approved by courts across the globe.
Court decisions have opened the door for the use of TAR and advanced analytics in e-discovery. A few jurists responsible for groundbreaking TAR decisions joined us for the annual judicial panel at Relativity Fest 2017. While the judges agreed that the workflow is increasingly accepted, the new question is how involved judges should be in protocol discussions.
With two separate and powerful workflows for technology-assisted review, you can amplify your team’s efforts on any project.
Create an approach that’s most effective for the unique needs of each project—whether it’s investigating the merits of a claim, sorting your data into key issues, or preparing evidence for litigation.
Let the analytics engine do the heavy lifting when it comes to prioritizing your review and preparing productions, so you can focus on reviewing only what’s relevant.
Use Assisted Review software workflows with other robust analytics features to quickly and effectively bring your story together.
“Analytics is on the rise throughout Australia as there is more litigation and regulatory action to make sure people are conducting reviews correctly. We have found active learning in Relativity has helped us seamlessly handle these complex reviews—and get the job done quickly.”
CRAIG MACAULAY, Executive Director, Forensic Technology
Active Learning in Technology-assisted Review: Relativity’s Approach to SVM and the Tech Behind It
Learn about the how the active learning workflow works, the technology behind it, and how to monitor and validate your projects.
Active Learning Overview
Learn more about how to use the new active learning workflow from our experts.
The Speed of Active Learning
Find out how active learning in Technology-assisted Review can help you get to the heart of your matter quickly.
Control Sets: Introducing Precision, Recall, and F1 in the Sample-Based Learning Workflow in Relativity Assisted Review
Learn the different ways to conduct and measure the accuracy of a sample-based learning project.