Body language. Facial expressions. Inflection. Demeanor.
Face-to-face interaction is data rich. Yes – those four examples are actually forms of data. We use what we glean from these, along with spoken words, to truly understand what someone we are speaking with or listening to means, their passion for the topic, their concern or excitement, and much more. The sentiment behind the words.
In the digital age, so much of our communication is electronic, not face-to-face. Emails, text messages, Facebook posts, Tweets, and various other platforms and apps help us quickly exchange ideas and “talk” when it’s convenient. Conventional wisdom tells us that sentiment is lost in translation when we cannot see or hear the people we communicate with. In a vacuum, that may be true. The reality is our electronic communications do not occur in a vacuum. A single text message exists as part of a macrocosm of all communications you have ever sent, including IMs, chats, emails, and social media posts.
Ask yourself what would happen if you took all available digital information for a single person, then analyzed it together. Most likely, we would see patterns and recognize changes in those patterns. We would see tendencies and variations. We would begin to discover nuances in a person’s communication style, such as where one tends to place emphasis, or phrases used with one person versus another.
Parsing an entire universe of information can help you analyze the underlying sentiment relayed in even a single text message. This is often referred to as sentiment analysis or opinion mining. Simply put, it’s the process of analyzing the attitude captured in texts and many other electronic communications. The science is already in use. One example is companies seeking to sell you products and services by working with Google, Facebook, and others to analyze your emails, posts, etc. They want to know what you think about, what interests you, and what you like, to better target their advertising.
Beyond commercial applications, sentiment analysis is becoming extremely important in litigation, where there is frequent discussion of "intent.” What impact does sentiment analysis have on that dialogue? To help litigators answer that question, they should work with an experienced forensics team to locate the right data sources and leverage the correct tools to assess meaningful sentiment in texts and other electronic communications.
From a scientific perspective, knowing what a person means in these exchanges can no longer be brushed aside. If we can successfully analyze sentiment in electronic communications, we may no longer need to read between the lines.
To access previous The Forensicators blog posts, please CLICK HERE.
Please email Mr. Williams at firstname.lastname@example.org to discuss sentiment analysis, digital forensics, collections, and other analytics issues, as well as how the experts at iDS can assist with all your legal technology needs on your next case or internal investigation.
The opinions reflected in this post are solely those of the author and do not necessarily represent the views of iDiscovery Solutions (iDS), nor any additional current or former employee of iDS. Moreover, any references to specific litigation or investigation work or findings are fact-specific.
Download Keys to Today’s Information Governance Landscape