WCC Smart Search & Match

In The Boardroom With...

Mr. Peter Went
CEO of WCC Smart Search & Match Thank you for joining us today, Peter. Please give us an overview of your background and a brief company history.

Peter Went: I have an extensive background in both software development and business consulting, and through that gained a great deal of expertise in relational database systems. I founded WCC Smart Search & Match in1996, because of a frustrating experience trying to find a new home to purchase. I realized at that time how ill-equipped existing databases were for performing certain types of searches. What I wanted was a technology that worked more like the human mind – something that created best-case options rather than looking for exact key word matches at the exclusion of everything else.

Out of that desire, WCC created ELISE, a high-performance search and match system that provides meaningful, ranked results across any type of data. ELISE is currently used in all types of arenas such as employment, ecommerce, enterprise content management, business intelligence and the security-sensitive identity matching field. Please give us an overview of ELISE, WCC’s Search & Match software solution.

Peter Went: ELISE is an extremely fast, scalable, vendor-neutral platform. In the identity matching market, ELISE uses multiple biographic and/or biometric criteria (known as multi-modal fusion) to find the most relevant matches. For instance, ELISE could search along facial scan, fingerprint, name and birth date to return the best possible match in sub-second response time. This technology works in border control, criminal investigations, disaster recovery and enrollment verification.

There is an interesting concept known as Information Magnetism that offers a wonderful explanation for how ELISE works. Information magnetism is the idea that just as we go out looking for certain types of information, information is also attracted to us. It might sound complicated, but we actually take it for granted in our everyday life. For instance, as a shopper you expect to be approached with products that might interest you. Imagine if you could only buy products you already knew about and specifically looked for.

Unfortunately this concept isn’t always used in traditional search. Typically, search engines only work in one direction – a user must search using precise parameters and even then they only get back limited options. ELISE works bi-directionally, meaning both the user and the information come together for the perfect match. In law enforcement for example, not only does a police officer search data using known criteria, but often data presents itself to the officer based on certain behaviors and characteristics of the person. It is this bi-directional data search that truly makes ELISE unique and the best choice for many situations. What are your core target markets and what are the market drivers in these areas?

Peter Went: ELISE provides a truly unique offering to a variety of market segments:

In the Employment arena, ELISE is used by staffing companies, job boards and departments of labor to match jobs and candidates worldwide. Unifying both structured and unstructured data, it provides meaningfully ranked results with data sources that run into millions of candidates and vacancies.

For the Enterprise Content Management market, ELISE is capable of finding and retrieving those ‘hidden gems’ that often get lost in peta-size archives. ELISE accommodates eDiscovery and Records Management with its ability to search multi-media archives including video and photo data.

Business Intelligence technology can go one step further using ELISE’s ‘guided search’ capabilities, transparency, and ‘bi-directionality’. ELISE offers real-time, ad-hoc analytics with the ability to search both structured and unstructured data.

Finally, Identity Matching is an aggressive new market for WCC. ELISE offers a vendor neutral platform for searching both biometrics and biographic data. Accuracy and speed-of-search mean customers have far less anxiety over security-related issues.

Part of what makes this such a hot market right now is that after 9/11, Identity Matching became a huge priority. Databases have grown exponentially as information on all international travelers is now collected and stored, rather than just a handful of suspicious ones. It goes without saying that there are significant issues around such things as properly identifying someone, dealing with those who might try and ‘spoof’ the system, and performance concerns for such a large amount of records. At the same time, there are significant privacy and security concerns that have to be addressed. These are some of the reasons WCC has taken such a prominent role in this space.
. We understand that you will be speaking at the I-Pira Biometrics Series May we have an overview of the key subjects you’ll be addressing?

Peter Went: I am very excited about the upcoming presentation, entitled “Multi-Biometrics Solutions: The 21st Century Identity Challenges. Guilty until Proven Innocent.” This is such a relevant topic for today’s world, and I’m anxious to be a part of the discussion. In the past, the majority of the world worked from the premise that you were innocent until proven guilty – that is, law enforcement looked for criminals only after a crime was committed. But there was a fundamental shift in that logic following 9/11, and now – with terrorism for example – we try to predict those people with terrorist tendencies and prevent them from committing a criminal act before they actually do it.

What that means is that databases are filled with many innocent people – in the U.S. for example they require data on anyone entering the country regardless of where they come from or their purpose. The information stored in these databases must be accessible in sub-second response time, the result must be conclusive – i.e. match or no match, no multiple choice, be able to deal with spoofing as a major concern. For this reason we are forced to apply the logic of multi-modal fusion and search for people using a variety of criteria. Security organizations realize that biometric data is more desirable than biographic, because it is more resistant to being manipulated. In addition, the more physical traits used to search the greater the chance that the match is truly accurate.

It’s not hard to understand the arguments for wanting greater amounts of information stored on individuals. There is a very valid counter-point to this however, that of privacy. This presentation will focus on the need for multimodal fusion and how we can solve both the need for greater information that is more easily and accurately accessed, while still protecting the privacy and security of our citizens. Thanks again for joining us today, Peter. Are there any other subjects you’d like to discuss?

Peter Went: I feel I would be remiss to overlook an exciting new concept concerning personal information and security. We know that there are multiple databases out there holding snippets of our personal information, and those will continue to grow as identity matching becomes more important. However, those databases aren’t very effective unless they can be combined and searched. For example, one database might contain your biographic data, another your facial scan and still a third your fingerprint. The smartest way to use that information would be to combine those records and properly search to find the best match. The fear of course is that if the information were to fall into the wrong hands, an individual’s privacy and security would be threatened.

The new concept is that of one-way scrambling. The idea is that you take information from each database and scramble it into unrecognizable code. A user can then search the centralized database and return a match, but no more than that; however that information would still be scrambled. Only by returning to the original owner of the database could that information be identified. It would allow us to take data from a multitude of sources and combine it into one record – a record that would hold all of your biographic and biometric data – yet it would be sufficiently scrambled so that no one would have the ability to understand it or de-scramble it. In this way it is completely meaningless. If I were to search my name Searching on first name ‘Peter’ in combination with some other traits on this database, it would first scramble it in the same way and then perform the search. On my end I would see that there was indeed a match for my search, however the information would still be scrambled and I would need to go back to the original data source to determine who the match was.

In addition to ensuring security, the concept of scrambling allows information to be kept in separate databases, bypassing the concern of adequate storage and performance requirements. It allows organizations to retain their proprietary information and circumvents political and business concerns.

This idea of one-way scrambling is still new, but at WCC these are the kinds of ideas that really get us excited. We’re in the driver seat of its invention and are hopeful that others, both vendors, universities and other stakeholders will join us as we all serve the industry together and have a vested interest in solving its chief concerns.