“I’m sick of it here. It’s just a matter of time before I take off.”
Imagine if John Doe, a key actor in your case or investigation, sent this message to his colleagues. It’s clear that John is displeased with his organization. Such an emotionally charged statement would provide perspective into John’s state of mind and help you find other people involved in the matter you’re looking into, right?
But let’s look at the linguistic pieces of his statement:” matter,” “time,” “sick,” “off.” At face value, no uniquely malicious words or concepts jump out. With only keyword or conceptual searching, it would be challenging to find this emotionally charged sentence among swarms of data, as well as the insights that come with it.
This is exactly the type of scenario where sentiment analysis can help accelerate reviews and uncover the stories in large data sets faster. Now available in RelativityOne, sentiment analysis detects positivity, negativity, anger, and desire across communications, providing visibility into the emotions buried deep in data.
With this new functionality, you can surface the most important information and expose narratives in documents that help tell the story behind your case or investigation. This AI-driven technology allows you to quickly identify the communications most likely to be relevant, requiring less manual review and speeding up discovery. Read on to learn about the different ways you can leverage sentiment analysis to navigate your data and expedite your path to insights.
Get to the Truth in Your Investigation
Investigating misconduct is complicated. It requires sifting through heaps of data and dealing with communication styles that vary across scenarios. Sentiment analysis levels the playing field by quickly identifying negative messages to help you determine whether individuals have engaged in problematic behaviors. This strategy can be used across many investigation types, such as:
- Employee Relations: In cases of harassment, wrongful termination, and employee discrimination, it’s the sentiment behind communications that demonstrate harmful intent. Sentiment analysis helps identify examples of potentially problematic conversations that investigators can review in context and judge accordingly.
- Detecting Fraud: When someone engages in illegal or unethical conduct, like embezzlement or withholding information that can affect stock prices, there are often conversations that hint at this behavior. Applying sentiment analysis across specific dates can divulge communications from entities that might have been “in the know” at the time of the unethical activity.
- Exfiltration: Unfortunately, sometimes data leaves an organization when it shouldn’t. Perhaps someone shares valuable IP material, or maybe an aggrieved employee departs abruptly with a book of business. Negative sentiment or anger in the communications leading up to that event can shed light on an individual’s motives and whether their actions were malicious.
To get the full picture surrounding an investigation, you need to determine not only what happened, but why it occurred, who was involved, and how they were feeling. With sentiment analysis in RelativityOne, investigators can discover the truth, no matter what situation they are dealing with.
New Ways to Prioritize Material
Prioritizing relevant material is always a key concern for review teams. As the volume of data continues to expand, it’s essential to find new and efficient ways to bring the most important information to the forefront. By quickly identifying powerful, emotionally driven documents, you can home in on information that is more likely to contain valuable findings.
Sentiment analysis populates multiple fields across documents. This gives review managers a variety of ways to prioritize material. For example, if there is a particular sentiment that is pertinent to a project, the Sentiment Distribution field allows users to filter by sentiment type and level. Or perhaps it’s of interest to find material with the highest total sentiment overall; in this case, sorting on Sentiment Richness highlights the highest scoring documents, which can then be isolated and batched out accordingly.
And with snippets of sentences directly viewable on the document list page, users can easily understand their sentiment results and determine the best way for the team to dive into their data.
Review Documents Before Production
Lawsuits always have two sides. While review teams spend much of their time trying to find the “smoking gun” in the documents received from opposing counsel, they are also responsible for managing documents that must be produced in return.
Sentiment analysis can help you understand what types of communications will be produced to opposing counsel before they go out the door. By viewing the documents with the strongest sentiment, teams can determine where they stand and catch anything an initial quality control review might have missed. Performing this analysis on your own organization’s or client’s set of documents might seem less obvious, but it can help you grasp what you are turning over, assess your level of risk, and ultimately impact the strategy you use in a case.
Don’t Hesitate to Be Creative!
RelativityOne customers have come up with numerous ways sentiment analysis can add value to their legal and investigative workflows. Some have applied sentiment analysis to court transcripts to identify judges’ positive or negative perceptions of certain issues. This approach can help determine which litigation strategies are more likely to be successful.
Others have gone even broader and started analyzing the sentiment of their call center transcripts. By understanding the types of communications that are more likely to result in positive conversations, they can train employees in accordance with those findings.
Leverage sentiment analysis to ensure that John Doe’s combative, yet relevant, statement doesn’t get lost within your data. Whether you are investigating misconduct, locating the most important documents to build facts in a case, or exploring a creative, out-of-the-box use case, seeing the emotional context across communications will help you discover the insights buried beneath the surface.
Want to learn more about sentiment analysis in RelativityOne? Talk with a member of our sales team to hear more about our latest AI functionality.
Graphics for this article were created by Sarah Vachlon.
Sarah Green is a product marketing manager at Relativity, leading go-to-marketing strategies for the company’s newest innovations and partnering closely with cross-functional teams on product launches and adoption activities.