BehaviorMatrix, an applied behavior analytics company, has received a patent for a system that classifies, measures, and creates models of elements that make up human emotions, shape their perceptions, and drive actions across the Web.
The now patented technology is based on the work of psychologist Robert Plutchik, who identified the eight primary human emotions: anger, anticipation, disgust, fear, joy, sadness, surprise, and trust. "Plutchik's wheel of emotions has been an important element of the formula that we've developed," Bill Thompson, CEO of BehaviorMatrix, says.
The approved patent serves as the foundation for BehaviorMatrix's emotional analytics platform, which will measure and deliver insight derived from emotional signals that permeate digital data such as social media posts and comments. While comparable solutions primarily rely on sentiment measurement, BehaviorMatrix's focus on emotion makes the solution unique.
"Looking at sentiment means determining whether something is positive or negative," according to Thompson. "Evaluating emotion is much more complex. We aim to get to the bottom of why a person feels what they feel. In other words, a sentiment would be 'I like apple pie,' whereas emotion would get down to 'why I like apple pie.' Does it trigger a childhood memory? Is it the taste? All these associations make up complex emotions," he explains.
BehaviorMatrix also offers a more holistic, computational approach to emotion analytics than traditional, keyword-based solutions. The company measures consumer data over an extended period of time, and leverages a combination of computational linguistics, big data algorithms, signal processing statistics, machine learning, quantum information theory, and crowd science.
"All of our analyses are quantitative. It's not just about looking at keywords and pinpointing whether they're positive or negative in a given interaction," Thompson says. "We use contextual NLP [natural language processing] technology and our other capabilities to get very calculated, accurate data. We don't just look at one set of feedback, for example. We collect and keep track of data for a period of years. Our customers know what we can do and they trust us to deliver high level data [to marketers]," he adds.
Armed with the emotional data, marketers can create a fuller picture of customers and be better equipped for designing personalized and targeted marketing campaigns. The technology, Thompson says, will help tackle some of the difficulties associated with gathering insight from unstructured data. The only type of data that BehaviorMatrix can't tackle yet is multimedia, which is poised to make up a significant part of online content soon. "The only thing we can't really touch yet is multimedia, meaning photos and video. We can't tell if there's a smile on someone's face in an image, for example," he says.
Though challenges remain, the solution is an innovative step toward deeper big data processing. "We're really the first in the industry to look at emotion expressed across all digital data, and this is a major advancement in the scientific world, Thompson says.