Your single source for new lessons on legal technology, e-discovery, and the people innovating behind the scenes.

Optimizing Index Training to Streamline Relativity Assisted Review Project Setup

Constantine Pappas

When computer-assisted review first gained popularity, we’d get calls from partners and clients interested in setting up a project. They didn’t know much about the process, but they knew enough to make them want to try it. These requests were often brought about by very pressing deadlines.

So we’d have a preliminary call where we’d explain what the project workflow entailed, and they’d be eager to know when they could start.

“Hold up,” we’d answer, “first we’ll need to build an analytics index.” Most of the time, one of two things would happen:  they’d have an index built using less than optimal settings for Assisted Review, or they would be unsure how to proceed.

We’d often provide our recipe, How to Create an Optimized Analytics Index, which has proved invaluable for folks who are just starting out. Our advice@kCura team is always happy to assist with these requests—but we also realize that the clock is ticking and clients are eager to get started.

Once the necessary saved searches were made, we’d instruct them how to spend a bit of time manually identifying repeated content and creating filters. And then, finally, we’d be ready to build the index. Some of the larger projects would take a few days for the build to complete, and if something was wrong or had been missed, they’d have to rebuild, which could take another few days.

The evolution of Relativity Assisted Review has introduced a great many features to ensure that a project will run more smoothly once it is underway. To name just a few, we have:

•    Application library installation
•    Automatic categorization set creation
•    Automatic report generation
•    Email notifications
•    Self-refreshing progress bars
•    Overturn reports with on-the-fly pivoting functionality
•    Integrated issue coding and weighting
•    Automatic batching

Once these issues were addressed, there were still setup hurdles to consider and address. Relativity 8.0 gave us automated repeated content identification—an extremely useful and much more thorough alternative to the manual approach. But the biggest setup hurdle still existed: the index itself.

All of this brings us to optimized indexing in Relativity 8.1. Admins no longer need to worry about file size, garbled text, and problematic file formats, which, if included in an index, could result in an undesirable impact on the index’s quality.

Instead, admins can simply enable the Optimize Training Set option and point it at a saved search. Relativity does the rest, and this ease of use, coupled with 8.1’s enhancements to analytics index build speeds, makes enormous strides in streamlining project setup for Assisted Review. In many cases, prep that once took days has been reduced to a few hours.

We’re not done making Assisted Review better and easier to use and, as in all things Relativity, we value your feedback. Please be sure to let us know how we’re doing.


Constantine Pappas is a licensed attorney with more than 15 years of legal experience. He has served as in-house counsel and managed both paper and electronic discovery for large-scale lawsuits and government investigations. As a member of Relativity’s customer success team, Constantine helps Relativity users with workflows for text analytics and computer-assisted review.

The latest insights, trends, and spotlights — directly to your inbox.

The Relativity Blog covers the latest in legal tech and compliance, professional development topics, and spotlights on the many bright minds in our space. Subscribe today to learn something new, stay ahead of emerging tech, and up-level your career.

Interested in being one of our authors? Learn more about how to contribute to The Relativity Blog.