Human head as a set of puzzles on the wooden background
PHOTO: Shutterstock

Last year when Sumit Agarwal set about to transfer his 300,000 airline travel points to his wife, he had an inkling that it would be an ordeal. Of course, he was right. The account had been dormant for a while plus his wife had a different last name so the transaction was triggering all kinds of security alerts. Ultimately Agarwal had to reset everything with his account to make the transfer. What made it all the more frustrating was that Agarwal, who is COO and co-founder of Shape Security, knew that it could have been so much easier if the airline had been making use of behavioral biometrics.

All it would have had to have done was flag the fact that both Agarwal and his wife have been on flights together. The probability of a thief (Agarwal) being on the same flight as his intended victim (Agarwal's wife) was practically zero, so ergo the transaction had to be what Agarwal said it was: an inter-family transfer. 

“Everyone thinks of behavioral biometrics as the speed at which someone types their first name,” Agarwal said. “That is a very narrow view of it. In fact it is your whole body moving around the world, your behavior in life, where you go in the morning for coffee, where you drive in the afternoon for work. All of that falls into the same category of data as the timing intervals between the keystrokes you enter on your phone or laptop.”

Related Article: What Customer-Centric Predictive Analytics Looks Like

What are Behavioral Biometrics

Behavioral biometrics systems monitor the user passively without any explicit gestures and try to deduce implicitly if it is the same user, explained Rajiv Dholakia, Nok Nok Labs’ vice president of Products. Some examples would be measuring different aspects of the mobile phone from the angle at which it’s held, to the user's gait while walking and the way the user types on the keyboard. These measures are all aggregated to create a "profile" for the user, he said.

It has been the explosion in the use of personal mobile devices and the Internet of Things that made new uses of behavioral biometrics possible as these devices provide the end points that gather data to authenticate users, said Nina Pineda, a consultant for PodM2M. When the user interacts with the device, “a process of continuous authentication is conducted enabling the biometric systems to capture not only the maximum amount of data to analyze the user, such as pressure, hand and finger movements, hand tremors, hand eye coordination, but also to detect anomalies in behavior,” she said.

A Customer Service Tool

Behavioral biometrics is often first thought of as a security strategy — and indeed it is being used by numerous banks and other companies in this way. But it is also increasingly being used as a customer service strategy or as a way to improve the customer experience. In these cases the wide range of actions that make up behavioral biometrics are proving to be very helpful. So, for example, by analyzing the speed at which a user name is being typed into a password setting for a retailer’s shopping cart, the store knows that the right user is signing in — let’s call her Sally — and not Sally’s teenage daughter Jane. Knowing that the site can respond appropriately with offers and marketing messages.

“Every interaction a person has with a brand produces data,” said Chris Connolly, vice president of Product Marketing at Genesys. “That behavioral data is being used for next-level fraud prevention, but also to give customers truly tailored and predictive experiences based upon what they are feeling or doing in real time.”

Related Article: What Does Evolutionary Psychology Have to Do With the Digital World?

5 Categories of Behavioral Biometrics

Scientists have classified behavioral biometrics into five categories based on where the data is collected, said Kim Smith, content marketing manager with GoodFirms. Below is how they break down.

  • Authorship Based Biometrics: A text, or painting or anything authored by a person is used to analyze the style and create a unique identity based on a wide spectrum of lines, strokes, handwriting, vocabulary and language.
  • Human-Computer Interaction Based Biometrics: The behavioral data regarding how a person interacts with computers and intelligent devices are used to create a unique identity of the person. This includes how an individual interacts with input devices such as computer mice, keyboards, control sticks as well as measuring advanced knowledge, skills, and intelligence while interacting with computers.
  • Indirect HCI based Biometrics: This is an indirect form of human-computer interaction, where data regarding common usage of computer for different purposes are measured, collected and analyzed. Call systems, web surfing, software usage, etc are all taken into consideration.
  • Motor Skills Based Biometrics: Muscle movements depend upon the proper functioning of the brain, skeleton, joints, and nervous system and thus, motor skills indirectly reflect the quality of functioning of such systems, which makes person identification and verification possible.
  • Purely Behavioral Biometrics: Humans use different strategies, skills, and knowledge while performing mentally demanding tasks. Purely behavioral biometrics quantifies such behavioral traits and makes successful identity verification a possibility.