How many times have you thought “I didn’t know that?” Maybe you’ve thought this after hearing the answers to Jeopardy questions (yes, even the kid version in some cases) or after finding out what makes a craft beer bitter. I have asked this question many times and thought it even more often. As the data we review expands from Giga-big to Peta-big, we need a faster way to find the important data – the needle in the haystack– the key facts for your case.
iCONECT XERA’s Advanced Analytics Clustering is the tool for that task.
Using Clustering to Discover What You Don’t Know (part 1)
When an eDiscovery review team is looking for the needles in the haystack, the color and size of the needles are often unknown, which makes finding them difficult. Data clustering helps organize the haystack, provides insight into the types of needles that are present, and makes the task of finding them less daunting.
In a nutshell, data clustering is the process of grouping documents using unsupervised machine learning of the document’s content and ideas without user input. The machine just learns what’s in the data and shows you.
iCONECT XERA’s Mass Cluster feature lets any user generate clusters with a few clicks. The result is a cluster tree with labeled categories and subcategories, which works just like folders
Clusters can also be visually represented using the iVIEW Mosaic Graph. Just like any other iCONECT XERA’s iVIEW graph, the Mosaic Graph can be refreshed after any search, which makes viewing clusters for a targeted custodian, date range, or keywords speedy and simple. #Easy eDiscovery
In the graph below, taken from our Sword Weaklegs case, you can see the clusters for one of our key custodians, Anders Franklin. The tiles in the Mosiac Graph (which represent clusters) can be used to filter, folder, or view the documents contained in the cluster. This makes it easy to know where to start reviewing and which documents can probably be ignored.
In this instance, the case is about athletes doping, so you might begin with the clusters in the red boxes and ignore the cluster with the yellow highlights. Before running any searches, you are in the dataset, learning about the case. You’ve got a head start on your investigation.
Next time I’ll look at how data clustering helps thwart sneaky actors.
To learn more about iCONECT XERA’s unique and easy to use Analytical Tools, check out our Analytic User YouTube channel.
April 28, 2017