by Joshua Whitney Allen

Personality, in the moment: IBM Watson and Chaotic Moon infer audience traits in real time

Published June 02, 2015


Last fall, as its Watson cognitive computing system grew in prominence in the culture’s fascination with instant technology, IBM opened the product’s gleaming new headquarters. The opening aimed to showcase Watson’s skillset, emphasize IBM’s commitment to its customers, and get people excited—which was easy to do, considering the opening was in the middle of New York City.

The IBM Watson Group is housed in a 12-story, elegantly modern building between Third and Fourth Avenues. It features a design studio and the Watson Experience Center, where IBM’s centerpiece technology demonstrates its powers through customer interactions and live demonstrations.

The opening of this center, Astor Place, last fall was a significant moment in IBM’s pivot toward a future of analytics, cloud, and cognitive computing. “The real estate move is the company’s largest relocation of researchers in its history,” Senior Vice President of the Watson Group, Mike Rhodin, said to New York Business Journal. “They’re putting together the researchers, sales and marketing teams, and client contacts designed to build up the Watson ecosystem, a collection of customers working on their own apps powered by Watson.”

The Astor Place debut, too, was a great moment to exhibit some of the capabilities of Watson, which carries a magnetism and appeal—almost a star quality, even. Chaotic Moon, an Austin, Texas-based development studio, designed and built a demo for a Chef Watson event planned at Astor Place in October of last year.

In just two months, Chaotic Moon created an interactive experience to demonstrate Watson’s potential. “What we’ve been doing is focusing on gaining insight on natural language streams of insights—Twitter—and leveraging Personality Insights as well as Concept Expansion and Message Resonance services,” says Marc Boudria, Solution Engineer at Chaotic Moon.

Key to the experience was Watson’s Personality Insights capability, which derives personality traits through linguistic analytics. The modeling extracts a set of personality and social traits from the way a person communicates through email, text, social media, and in this case, Twitter.

The team had screens set up that showed baseline emotional resonance based on the guests’ Twitter handles. “We were able to add new data while people were tweeting to a hashtag [set up during the event],” explains Boudria. “Using Watson Personality Insights solution, we analyzed the audience reaction and displayed the real-time results.

“Because we had Twitter handles, by using Concept Expansion and Relationship Extraction services, we were [later] able to show the relationship of everybody in the room.”

The exercise was a real-time immersion in audience analytics, tailored messaging, and response metrics. It was a great display of Watson’s acumen at text analysis, and the Astor Place housewarming also explored cognitive computing’s broad abilities to amplify and contextualize modern communications. Watson’s Concept Expansion, Message Resonance, and Relationship Extraction are tactics that augment marketing into a refined and measurable practice.

The IBM Watson Personality Insights service uses linguistic analytics to infer cognitive and social characteristics from user communications. Digital communications could infer models and patterns of personality characteristics, basic human values, and needs derived from text. “We are not calling that profiling; we create a portrait based on analyzed texts and infer characteristics that describe a person in a new, more meaningful way,” says Maria N. Schwenger, IBM Senior Software Engineer.

With these inferences, marketers can shape messages that are tailored to the tastes and outlooks of their customers. To understand nuance is a priceless skill in the exchange between business and consumer, and Watson’s Concept Expansion is intended to capture meaning in the culture’s ongoing dialogue. This service analyzes text and interprets its meaning based on usage in other similar contexts. As noted on IBM’s site, it can interpret “The Big Apple” as “New York City.” Users can create a dictionary of related words and concepts, euphemisms, slang, colloquialisms or otherwise unclear phrases can be better analyzed.

As Boudria explains, the Chaotic Moon use case prioritized Message Resonance. This service analyzes draft content and scores how well it is likely to be received by a specific target audience.

Watson’s Relationship Extraction capability parses sentences into their various components; it scans for links between the components and allows users or analytics engines to grasp the meaning of copy and documents.

“Based on years of research conducted in IBM Research Labs in linguistic analysis and literature from psychology theory,” explains Schwenger, “[it] offers a set of core analytics to infer the personality characteristics, intrinsic needs, and values of individuals from communications that the user makes available via mediums such as email, text messages, social media, forum/blog posts, and more.

“These insights help clients to understand their customers’ preferences and improve customer satisfaction by anticipating customer needs and recommending future actions.”

Context is everything, it is said, and Watson’s approach to context has led the technology into several active partnerships throughout industry. The Watson Engagement Advisor, a SaaS product that assesses natural language with contextualized, evidence-based reasoning, is at the center of a collaboration between USAA and IBM.

According to IBM, USAA teamed up with IBM Watson to provide a one-stop shop for veterans seeking to manage job hunts, benefits, and life changes. To offer the most relevant content for users, Watson analyzed more than 3,000 documents on topics exclusive to military transitions, enabling members to ask and receive answers to separation-related questions.

With the economy so often a series of choices—which product to buy, which mortgage to select—inferences on one’s mood would be valuable in making decisions. Life is a series of choices and as the digital age moves faster and faster, IBM intends Watson to help. Choosing an entrée at a restaurant, picking a hotel for vacation, deciding which mutual fund best matches one’s comfort with risk—cognitive computing and Watson is meant to simplify these tasks.



No one has commented on this item.