2 min read

Office 365 E5 Enhancement Part 1: Advanced eDiscovery

Featured Image

Over the next few weeks, we’ll be reviewing the exciting enhancements in Office 365 E5, such as security improvements and data loss prevention (DLP), but first let’s review Advanced eDiscovery, formerly known as Equivio Analytics. 

There is no doubt about it, business intelligence (BI) is extremely important ediscovery-image.pngtoday and with the volume of data that confronts us every day, it’s for good reason. It’s much easier to store data these days, but making that data mean something to you and your business is the challenge and one that Office 365 simplifies with the new functionality in Advanced eDiscovery.

Advanced eDiscovery, or electronic discovery, uses advanced text analytics and predictive coding to perform multi-dimensional analyses of data collections. Technical jargon aside, access and analysis of stored data can help you save time and money in many business cases. It’s a form of business intelligence that allows a company, to search, identify, and deliver documents and data needed for regulatory compliance, internal investigations or possible litigation.

Microsoft’s Mechanics YouTube channel presents a great example of just how powerful this functionality can be. In this video, a query is run, analyzing data from Exchange, SharePoint, and Skype within Microsoft Office 365. Not only does it include the data from the online systems, it also includes data from outside sources such as Facebook and Twitter. After the data extraction, users can narrow the data set down by excluding duplicates based on the set algorithm. Email Threading also helps to identify the unique messages in an email thread so you can focus on only the new information in each message. 

 After running the initial analysis, the operator is brought through a predictive coding step so the system can be trained on what you’re looking for, marking certain emails or data as relevant or not relevant to the particular case. Then the system reports on whether the algorithm is stable, making sure that enough data has been reviewed manually. This process validates the machine learning steps against a “sample set” which then validates the accuracy of the process, demonstrating the cost effectiveness of manually reviewing a sample set of documents, versus allowing the machine learning algorithm to review a much larger set of data.

Themes help speed up this process by allowing users to get valuable insight about data beyond just keyword search statistics. By grouping related documents, users can look at the documents in context. When using themes, you can view the related themes for a set of documents, determine any overlap, and then identify cross-sections of related data.

After the data analysis and eDiscovery investigation, users can then export the data. The export package includes a CSV file that contains the properties from the exported content and analytics metadata. This export package can then be imported to an eDiscovery review application.

Check in next time for more E5 enhancements and let us know if you have any questions in the meantime…we’re here to help!