Questions to Ask Yourself:
1. Are you relying on your service provider’s analytic software alone?
2. Do you know how eDiscovery software classifies your data?
3. Do you know how eDiscovery software organizes your data?
4. Do you know the science behind classifying and organizing your data? We do.
Every eDiscovery company will tell you they have the latest and greatest software to perform all sorts of miraculous feats. We do too! You can’t get away from software – the volumes of data to be reviewed dictate that you have to have the software necessary to perform these activities. The difference is that we actually understand how these techniques work and we’re happy to discuss the appropriateness of an application or modifications customized to your needs.
Using Viewpoint, we have many techniques we can apply to systematize and streamline your document review. You can think about these techniques as either structuring your records for review, or cleaning the records to get minimize duplication.
Structuring records involves creating a system for classifying documents by type and content – like creating a pigeonhole filing system. Depending on what you want to know, there are several ways to do this. The best known is predictive coding, which involves a review (by a person) of a small sample of documents to codify them into categories. Using these categories, the remaining documents can be codified in a first pass. A sample of these are selected, reviewed, and reassigned to categories with the possibility of new categories added to the first set. The process can continue until it converges to a final set of categories that you find most helpful. This lets you determine which documents ranked according to how likely they are to be responsive in your case, and where to expend reviewer resources.
You can also conduct content analysis to capture specific concepts and cluster documents together according to the concepts they address. A map of clusters can show you how frequently some concepts arise, and how do they overlap with other concepts, helping you identify important topics and again let you prioritize review.
The simplest organizational tool you can use is searching for and finding particular terms. Like an index at the back of a textbook, you can list topics of interest and see visually how frequently certain terms are found.
Structuring records as described above involves creating categories and differentiating. But some records are duplicative and you’d prefer not to review them if you have already touched them once. We can eliminate exact duplicates from the databases being reviewed and find near-duplicates so that review is limited either to single exemplar, or to comparison between near duplicates to determine if there is an interesting reason why the document changed. The same is true with emails and email threads. Messages and users can be grouped to find unique content and sets of communicating users. Email relationship analysis expands on this capability to find the frequency of communications and communications within a ring of participants.