Data is the basic building block of analytical exploration. It tells a story—in an infinite amount of ways. But how do we read the story? How do we choose what chapters to focus on and where to look for clues?

For e-discovery professionals trying to take a massive heap of data and find the 25 documents that are important, this can be an especially painful challenge. The ESI in a matter come in so many different forms and are related in any number of ways, seeing meaning can be near impossible with existing data visualization methods.

The Problem With Basic Data Visualization

Emails, pictures, audio files—all of it is data. But the way we organize it and interact with it can either give us limited or unlimited insight. How we understand data is impacted by what we see. Visual cues quickly allow our brands to focus on one thing or filter out other things.

We're in a never-ending cycle of trying to interpret data and glean information, but this requires us to see the data from different perspectives. While advances in technology have resulted in crucial gains related to speed and intake volume, human intelligence is still inarguably an essential element of understanding the intelligence behind the data.

Luckily, we have this amazing thing called vision. Out of all our senses, it collects and processes the most amount of data, giving us a better view—literally—of the world around us. Studies in UX and how various design aspects are processed by our brain are also allowing us to better express information in a way that unlocks new understanding.

Leveraging both technology and our own power to understand can help reveal data intelligence faster and with more accuracy than any AI-only or human-only process. AI can be enhanced with successful data visualization. For example, AI training models can be created by selecting a group of items that were determined to be relational based on manual data visualization. Or, basic AI insights can be presented for final review and assessment in a way that is visually meaningful to the users.

What Does Successful Data Visualization Look Like?

The ultimate goal of data visualization is this: to gain insight.

Seeing the data in a way that imparts meaning is only half of the equation. The other half is the ability to navigate into the data being shown. If Google Maps didn't have a zoom feature, for example, it would be essentially useless.

Navigating the data means the ability to dive in deeper and select groups of data items or relational dimensions, or focus on particular sections of the data at any given moment. To choose some things and filter other things out. Like turning the knobs on a microscope or a telescope until the intelligence comes into focus.

Let's take a look at a few examples of different ways to present data that could reveal new insight.

• Thread arc.

Now imagine seeing parent/child relationship (arcs) + chronology (x-axis) + variable A (color) + variable B (distance between dots) all at once. If you have a long list of email threads, with a thread arc, you can cross reference 10 to 20 at any given moment just by using your eyes.

• Treemap.

This represents a dataset of parent/children relations. So, for example, you have tags on certain folders and documents, that can be used to create a quantitative relationship. That score can now be used to show the folders with the most relevant documents as larger as compared to smaller folders which would represent less of a relational similarity. Size can represent whatever the score is, for example by the amount of bytes each folder takes up, in this case size is just the relation.

• Circle-packing.

This visualizes the same, hierarchical data structure, but shows how items are nested with a different positioning algorithm.

While pie charts and bar graphs have their place—dashboards and as elements of a larger, more detailed chart, for example—it is helpful when working with big data to break away from relying solely on these simple constructs otherwise, we are severely limiting what kind of insight we can gain from the data.

The vast majority of e-discovery solutions force you to choose from a limited selection of typical charts, but real insight and efficiency requires flexibility. The flexibility to use more than just an x-axis and a y-axis. The flexibility to use color, texture, shape, orientation, size and nesting to indicate relationships between your data—and literally see meaning. The flexibility to discover something different.

As data grows exponentially, and firms and corporate teams are required to gain better insight from more data faster, the freedom to access custom visualizations for unique use cases during the discovery process will change the way we do e-discovery.

Jonathan Snyder serves as UI/UX lead for LegalRadius, a ONE Discovery software solution, where he manages all conceptualization and design within the software application.