Biometric Dynamic Profiling Techniques

The Evolution of Fingerprint Biometrics

Fingerprint biometrics technology has been an invaluable tool in the criminal justice system for many years, and has been used by thousands of professionals to solve crimes and confirm the identity of suspects.  The FBI’s use of fingerprint technology in particular is well-documented.  Containing over 45 million records, it is easily the world’s largest, and they recently announced a desire to integrate their national database with criminal database records from around the world.  This would enable their organization to improve capture rates and more closely monitor fugitives globally.  They have also recently made public their plan to develop what they are calling the “Next Generation Identification System”.  If completed, will become the most comprehensive biometric database ever developed.  Because the genesis of biometric technology began in criminal tracking, it tends to receive the most publicity and to this day is the discipline to which the average citizen associates fingerprint technology.  However, biometric fingerprint technology is rapidly emerging in the private sector as a way for organizations worldwide to improve the identification and management of personnel and membership in a variety of industries and environments.  

The Driving Forces Behind Biometrics

Biometric technology firms develop fingerprint biometric alternatives for companies that deploy a unique identification system that controls, manages, or tracks access to their database applications.  Historically, these ID systems have been managed by using barcode scanners, card readers, user id/password techniques, etc. The integration of biometric technology can benefit organizations in different ways, depending on where and how the system is deployed.

For example, firms that design workforce management/time and attendance solutions have integrated fingerprint technology with their software that enables customers to more closely track and control their labor costs by eliminating “buddy punching” (employees clocking in for one another).  Membership management solution providers are using fingerprint technology to help clients control access to their fitness centers.  Prisons use the technology to improve security by tracking prisoner movements and controlling access to restricted areas, and schools are using the system to manage lunch-line point of sale processes.  The use of fingerprint technology has helped companies and organizations cut costs, improve the accuracy of data being collected, prevent unauthorized access to sensitive information, and more.  There are often objections that arise when the idea of integrating a fingerprint technology solution is discussed, which are typically centered around concerns of privacy and identity protection. 

The Fallacy of Privacy Risk

You can’t have a discussion about fingerprint biometrics without discussing the issue of privacy and the protection of biometric information.  Because of its common association with the criminal justice system, fingerprint technology in the private sector is often misunderstood.  According to Michael Trader, President of M2SYS Technology, a leading biometrics research and development company based in Atlanta, GA, “Sophisticated biometric fingerprint systems do not endanger privacy protection in any way because they do not store optical images of prints.  Although an optical image is captured during the initial registration process, it is immediately converted to a digital file and the image is destroyed.  All that remains is what is referred to as a ‘binary template’, which is of no use outside of the database, and cannot be reverse engineered to recreate the optical image of a fingerprint.  In essence, the fingerprint template is no different than the numbers on a barcode reader.  The great advantage, however, is that unlike ID cards, user names and passwords, or other forms of identification, the user can’t forget their fingerprint, nor can it be stolen from him.”  Adds Mizan Rahman, CEO and Chief Scientist of M2SYS, “Biometric identity management technology is based on who you are, rather than your possession of an ID card or knowledge of a password. Biometrics is not meant to entirely replace the existing methods of identification, but to build on them in order to strengthen and enhance security benefits.” 

False Rejects and False Accepts

Understanding the limits of fingerprint technology is just as critical to understanding the benefits.  Depending on the environment in which a fingerprint solution is deployed, expectations and performance may vary.  For example, in a 1:1 (one to one) database environment like the one used by law enforcement agencies to compare a latent print against a potential match, security and accuracy is obviously critical, convenience and speed is less important.  When attempting to confirm the identity of a criminal subject, for example, it is more important that the system returns an exact match, regardless of the time it takes. In a 1:N (one to many) environment, speed and efficiency reign supreme.  For example, if 200 children are attempting to use the fingerprint system to check-in to an after-school program, it is more important that the system works efficiently and at a high rate of speed.  In either environment, reject rates can surface, and are referred to as either false reject or false accept rates.

A false accept occurs when a user who has not previously been registered in the system (and therefore has no biometric template captured) receives either authorization to access the system or is granted identity verification.  It can also occur when the system falsely identifies the user as a different user in the database.  The probability of a false accept occurring is known as the false accept rate (FAR).  False accepts are generally quite rare in 1:1 systems that employ a respected biometric matching algorithm.

A false reject (or false negative) occurs when the system fails to identify a successfully enrolled user.  In other words, although a biometric template exists for the user attempting to be identified, the system returns a “no match found” result, and user authentication is denied. If the user’s biometric fingerprint template image is slightly misaligned during the initial capture, they are forced to repeat the misalignment during each subsequent scan, or face being falsely rejected.  The probability of a false reject occurring is known as the false reject rate (FRR).  An increase in the FRR will result in more users being required to produce multiple scans, and in application environments such as workforce management and point-of-sale where time and efficiency is of the essence, an increase in the FRR can result in lost productivity, decreased efficiency of use, and an overall increase in frustration over the system’s usage.  

“Developing solutions that reduce the false reject rate is critical to the overall success of any biometrics system”, says Trader.  “Not only do false rejects increase the user’s frustration level, in certain environments such as Point of Sale and Time and Attendance, it can adversely affect productivity and ultimately profitability.”  To combat this problem, biometric systems must be programmed to “learn” from the samples and ultimately make a decision that is based on an aggregate set of fingerprint samples, as opposed to creating a match using a single sample comparison. 

Next Generation of Fingerprint Matching

In order to combat the problems associated with false rejects, M2SYS Technology has recently developed an “artificial intelligence” component for its fingerprint identification systems.  Referred to as Bio-AI™, this component uses subsequent login data files to build a much larger biometric template file for each individual. By utilizing this functionality, users of the fingerprint system have been able to dramatically reduce the false reject rate and improve the system’s throughput.  This is accomplished by eliminating the need for users to perfectly align their finger during identification scans, improving usability and increasing the fingerprint matching likelihood.  “False rejects are a common but very serious issue across the entire biometrics industry”, states Rahman. “Undeservedly being unable to clock in for work or being denied the ability to pick up your child from daycare can have serious consequences. With Bio-AI, M2SYS is again solving a vital issue in fingerprint biometrics and reasserting its role as a biometric technology innovator.”

About M2SYS Technology

M2SYS Technology, www.m2sys.com, is a recognized industry leader in fingerprint identity management technology, delivering a wide range of turnkey fingerprint software solutions for businesses and consumers.  M2SYS was awarded the prestigious Frost and Sullivan Biometric Technology Innovation of the Year award in 2007.