Thought Leadership and Industry Trends
Useful Relativity Features You May Have Never Heard Of: Part Two – Review and Beyond
In part one of this series, we discussed some of the features in Relativity that can be used to investigate document sets before the processing and review stage of discovery. The following article explores two Relativity technologies that can assist users when they are reviewing documents and preparing them for litigation.
The key advantage to using any form of legal technology is reducing the number of time-consuming manual steps need to be taken in order to accomplish a goal. The tools built into the Relativity platform help gain users advantages and automate aspects of the time-consuming discovery process.
Continuous Active Learning
Continuous Active Learning (CAL) is a form of Technology Assisted Review (TAR) that leverages machine learning to identify and prioritise the most likely relevant documents quickly and efficiently. This allows a small case team to work through a large amount of documents quickly and efficiently by constantly training the system as to what constitutes a relevant document.
In the past, Relativity Assisted Review (RAR) consisted of reviewing a sample of documents, and then using this sample to train the system to make coding decisions. Training the system was often time consuming and caused downtime in the review, as additional review and training could not be undertaken until the system had finished learning from the previous round. Once the system completed its training, a new set of documents were made available to be reviewed and the results were analysed to calculate the quality of the review. This process would repeat itself with each round, training the system further until the quality of the review met with an agreed-upon level of accuracy.
CAL differs in this approach by aiming to provide a continuous queue of documents for review. As the documents are reviewed, the system trains itself in real time. As the system learns the difference between relevant and not-relevant documents, the system pushes the most likely relevant documents to the top of the queue. This process continues until all relevant documents have been reviewed and only not-relevant documents remain.
The use of CAL allows small review teams to quickly code a large case of documents quickly, defensibly, and cost effectively. The “training round” method of TAR was best suited to large cases, due to the amount of sample documents which required review in order to train the system to a defensible level. CAL opens up the usage of TAR to much smaller cases, alleviating the need for using keywords, which in itself is a more defensible approach. The techniques listed above such as concept searching, find similar documents, and clustering can be used alongside CAL in order to jump start your review. A group of highly relevant documents can be coded in advance and then used to train the system on relevance quickly and effectively, insuring that the review team is looking at relevant documents immediately and the system can begin learning and pushing similar documents to the top of the queue.
As a Relativity project manager working outside of a law firm, it is often near the end of a case when a client working within a law firm sends me a large Word document. This Word document contains a table listing the Relativity document numbers (sometimes containing typographical errors, inconsistent formatting, or unusable content) and a description of the content of each document. Although this document is interesting and a lot of work has gone into its creation, it does not help a Realtivity case manager to perform any request in a way that correlates to Relativity, in turn creating work that is both expensive and time consuming. This document also requires manual formatting and updating as new documents are located or as documents are removed.
Case Dynamics (previously Fact Manager) is a Relativity Application that can be used to document the key issues, people, organisations, dates and corresponding documents in your case. When accessing the Case Dynamics application within Relativity, documents are coded to facts in the same way as a document is coded to relevance. This coding can be completed during an initial first stage review, in conjunction with a second level review or as a standalone review for trial bundle preparation.
Once documents are coded for facts, people, and organisations, reports can be viewed within Relativity with direct hyperlinks to the documents. Not only does this application track all of the information that is manually added to cumbersome Word documents, but it does so quickly and efficiently.
Case Dynamics also provides more advanced reporting than that of the manual and cumbersome Word document, allowing the user to export hyperlinked reports containing PDF copies of the referenced documents. These reports can then be viewed offline, shared internally, or taken away for custodian interviews and offsite trial preparation.
Alongside offline reporting, timeline reporting can also be utilised to visually display a colour coded timeline. This timeline can be created using a range of criteria i.e. limited to specific people, organisations or facts and provides an easy to digest visual representation of how a case progressed. Alongside the visual representation, the timeline can be used to navigate the facts and documents within Relativity or exported to an offline report as a further resource for trial preparation.
The five technologies discussed are just a few of the less well known, overlooked and powerful features that exist within Relativity and RelativityOne. To discuss how CDS can assist you with leveraging these and other technologies, contact us for a consultation.
About the Author