HaystackID® Acquires Business Intelligence Associates, Inc.Read More

eDiscovery Analytics Tools: How to Choose and Use Wisely

eDiscovery Analytics Tools: How to Choose and Use Wisely

Do I really need all these eDiscovery Analytics tools?

To many attorneys and lit tech personnel, eDiscovery analytics tools may feel a bit like KitchenAid mixer attachments. Users can’t work them, they don’t have time to learn how, and new gadgets hitting the market only add to their stress. Often, the result is a lonely arsenal of sophisticated tools (the spiralizer, grinder, or dough hook of eDiscovery analytics) that go unused while litigation teams dutifully review one PDF at a time, the equivalent of stirring brownie batter with a fork.

The stigma surrounding eDiscovery analytics tools—too expensive, too complicated, too big for small cases, etc. —is unfortunate but far from groundless. The technologies are highly sophisticated, constantly evolving, and require constant upkeep and user training. In the hands of trained experts, however, these tools can save you significant time and money over the lifespan of your case. Considering that saving time and money is the overarching purpose behind adding eDiscovery analytics tools to your workflows in the first place, why not put those tools to proper use in ways that best achieve that goal?

Many of us refer to “analytics” as a blanket concept in our everyday on-goings, and indeed, analytics are responsible for much of our daily intake and output of tailored digital information. The reality is that there are many, many ways to use analytics capabilities. Just as Google, Amazon and Netflix use different types of analytics for different purposes (subway maps versus shopping ads versus movie suggestions), owners of an eDiscovery analytics toolbox can put those technologies to work for a wide variety of very specific tasks, especially with expert assistance from eDiscovery handymen like our team at BIA.

A Brief Intro to Primary eDiscovery Analytics Tools

We often say in eDiscovery that he who holds the data holds the power. For attorneys and lit support personnel, however, the power and possibility of data is only as strong as the analytics—the mining tools that extract the relevant information “gold” from your mountain of data. Analytics can and should be your trusted companion throughout every step of the EDRM—from litigation holds to case preparation to document review and production. At the front end of the eDiscovery lifecycle as well, there are tools to assist with early case assessment, determining the scope of the case, and predicting budgets.

At BIA, the analytics tools we use and discuss the most are those that help to narrow a data population and assist with document review. Here are some of our favorites:

Textual Analytics Tools

  • Technology-assisted review (TAR) 
    • 1.0 is a predictive coding where the analytics technology uses the coding from one set of documents and applies it to the rest of the docs.
    • 2.0 is a machine learning engine with continuous multimodal learning capabilities that renew the coding after every doc, continually teaching the machine how to review more accurately with each new coding call.
  • Emotional intelligence and heat-mapping
    • Offers coding calls based on the depicted emotional state of the person writing an email. What can we learn about custodians by the language they use? Or the punctuation they use? The time of day they’re writing?
  • Concept searching
    • Takes concepts (which can be as simple as a word or a phrase) and offers other search suggestions that the user may not have considered.
  • A cluster wheel that organizes documents by elements and relationships
    • Users can follow the wheel’s outward extension for more specific areas and ultimately single docs to be reviewed or coded.

Relational Analytics Tools

  • Identification of near-duplicates and exact or textual duplicate identification.
    • Instead of relying solely on hash values, near-duplicates are determined using a combo of text and metadata as well, thus reducing time wasted reviewing dupes or near-dupes.
  • Email threading
    • Sets aside redundant material (replies, copies, forwards) and neatly organizes long e-mail chains so that they are reader and reviewer-friendly—think Marie Kondo for your e-mails.
  • Communication maps
    • Shows patterns of conversations. Who’s having the conversations, how often, what are they talking about, and how invested are they?

Analytics Tools Developed Specifically for the Legal Industry

  • Name normalization
    • Consolidates the various ways a person may be identified (e.g., ties one person to 3 different email accounts).
  • Redaction
    • Removes private or sensitive information for document production.
  • Personal data (PII) detection
    • Highlights social security numbers, phone numbers, health records, etc., for redaction.
  • Language identification
  • Metadata remediation
    • Takes a PDF and pulls the “from,” “to,” “CC” and “date sent” data and puts it into a field so that it is searchable. This is especially helpful in cases where processing was sub-par and you only received only PDFs with no metadata other than who created the PDF.
  • Metadata enrichment
    • Categorizes documents even further in ways that can speed up review. It can identify very large families, for instance—a big time-saver if you are doing rolling productions and need to get docs out quickly.  
    • Generates a report showing the number of recipients of an email. Did the email go to 1-5 people? 6-20? More than 20? I like this tool for privileged review because if an email went to more than 10 recipients, I know it’s probably not privileged. I can create a search and eliminate docs that way, rather than reviewing one email at a time.

Other Miscellaneous Tools

  • Photo descriptions
    • Looks at images in the background and describes that image with text. For example, if the tool reviews a picture of a construction site and detects a tree, it will enter the word “tree” into a field so that it’s searchable.
  • Facial recognition
    • Analyzes and pulls data from a video. If the tool recognizes a person in the video, it can pull up additional data on that person. And it’s not just faces—for instance, if it recognizes a company logo on a NASCAR race car, it can open the link to that company.
  • Automatic transcriptions of audio and video files
    • Adds transcription of the text to the metadata of the files.
  • Social media investigations
  • Data visualizations
    • Offer different visual modes so users can choose the one most conducive to accurate review of a particular set of data.

Do Try This at Home—With Our Help.

The list above is only a sampling of the analytics tools in our BIA tool belt and not a comprehensive one by any means. Learning and running these tools to their maximum capacity is a full-time job. If you are an attorney or lit tech professional trying to run eDiscovery analytics tools yourself, you are not alone in feeling overwhelmed, ill-equipped, or ready to throw your hands in the air and run for the hills. To garner the most time and cost savings with your analytics tools, your best bet is to work with experts who not only know how to use the tools, but who have the knowledge and experience to choose which tools will most benefit your individual case and the stages therein. Acquiring the tools is only half the battle. Without a thorough understanding of the tools’ intended use and capabilities, you’re less likely to see the time and money savings that your investment could bring.

Take email threading, for example. Many people use this popular tool primarily as a document organizer. Certainly, there is immense value to having your docs well-organized, all in the same chain, ordered consecutively, etc. But the real game-changer here is the threading tool’s ability to identify unique documents. (An example of a unique doc is the very last email in a thread, i.e., the only one that contains all other emails before it.) If you do email threading alone, without the “unique” or “inclusive” feature turned on, you still have to look at every document and every individual email in the thread. Many users miss this feature and end up reviewing an average of 20% more documents than they needed to. Enlisting our team could help those users save or effectively redirect that 20% time and energy.

Another tool I see clients commonly misuse or under-utilize is the text duplicate (or exact duplicate) identification tool. They will use it to access other duplicate documents but neglect to turn on the propagation feature which automatically applies coding across a relational group. If you are using text duplicate ID but not also using propagation to copy the tag to every other duplicate responsive, once again you are creating—unnecessarily—more work for yourself and your team.

Machines Need People

While the world keeps producing data and companies continue to enter litigation, it’s safe to presume that new analytics technologies will emerge while the old ones get updated, augmented, and overhauled. The job of staying on top of the tools and their changes is not getting any easier for the attorneys and lit tech staff managing document reviews. The eDiscovery analytics tools we’ve discussed here (and many others that we haven’t covered) were created and designed to assist human reviewers and improve review speed and accuracy. But to achieve the utmost in speed, efficiency, and accuracy, even highly sophisticated machines and technologies still need (for now, anyway) competent humans to learn and run them.

At BIA we have the analytics tools and the humans to run them. Our teams began mastering and even building tools back in 2002, and we never stopped. Our full-time managed review team of attorneys are trained, certified, and ready at the helm to handle any task or query. We can assess your case needs and demo new tools, or we can help you dust off any eDiscovery Analytics tools you already have and get them working for you.

Daunted by your eDiscovery analytics tools? Let BIA be your helping hand. Reach out today.

Aaron Boone, Esq.

Aaron Boone, Esq.

Aaron Boone is part of BIA’s Litigation Technology Expert team. With over ten years of experience working in litigation support, he has a vast knowledge of the various industry-standard analytical tools such as Brainspace, Relativity Analytics, NexLP, and Matter Analytics. Aaron holds a JD from Western Michigan University’s Thomas M. Cooley Law School and is certified as a Brainspace Administrator and Milyli Blackout Administrator.