Why idMatch™ ?

Am I Ms. "Sarah Abramson" or Ms. "SSN# xxx-xx-xxxx"?

How would you like to be called? Obviously by your personal name, that is why we have personal names. We always like to be associated with our personal name, rather than some machine generated random identification sequences. This is the fact that the businesses understand, they understand that the personal names are critical business assets. Corporations track most of the business transactions by the customer's personal names. The best and the worst revenue generating customers are identified by personal names rather than numbers. This is the case of any revenue generating industry. The same kind of problems exists with the government security agencies or any other institution where personal details scanning are of utmost importance for the business to run profitably. A few industries to name are Banking, Import/Export industry etc.

Some facts about Names

Let's start with a few facts about names:

• Names are not unique
• When said, written, entered into systems, are subject to considerable variations
• These variations/errors cannot be practically avoided (Cultural, Linguistic, Ethnic boundaries)
• Even when the source data is perfect, when searching, the user may be searching for wrong spelling

Knowing these facts about personal names, now comes the problem of finding the right identity or person in the database. For some industries failing to find the person/entity in database and creating new account/entries for the entity may be only one more entry in the database of entities, however for most of the businesses or government organizations it is not so. For all the industries, being unable to find the entities in the database may lead revenue loss, one way or the other. The reasons being many and all of them standing correct in one situation or the other, to list, a few of them are:

• Duplicate entries may cost the handling/maintenance cost
• Duplicate entries may increase the postal charges incurred to send information to the entity due to the duplicated data
• If the match being done is against a compliance list like OFAC list, unable to identify a blacklisted entity may lead to heavy business loss

How do we solve this problem?

There are multiple ways of solving the problems with varying degrees of relevance. There are many algorithms that tend to pose partial solutions to name search problems however all of them are unable to provide the accuracy level that is expected today. These solutions/algorithms were developed long back, some of them like SOUNDEX dating back to early 1900's and were relevant to the name data that existed in that period of time. Since then, businesses have come a long way, seeing huge variations in name and the way names are treated for the transactions. The laws have changed, so has the naming conventions and the boundaries for doing the business. Today as the businesses are going global, these algorithms, which are basically English oriented, do not scale up to the business needs, in its relevance level. To solve these problems we need an advanced name search software tools. Even today most of the organizations rely on the poor name search technology like SOUNDEX developed in early 1900's. The SOUNDEX algorithm ignores common personal name variations based on culture, language, spellings etc. Studies prove that SOUNDEX fails to return the correct match that exists in the database, however at the same time it returns bulk of unrelated false positive matches. This leads to increasing the risk in business making the business prone to working with customers linked to criminals, fraudulent, or terrorist activities. Many studies have been done on SOUNDEX, and a few of the results are:

• Only 33% of the matches that would be returned by Soundex would be correct. Even more significant was the finding that fully 25% of correct matches would fail to be discovered by Soundex. (Alan Stanier, September 1990, Computers in Genealogy, Vol. 3, No. 7)
• Only 36.37% of Soundex returns were correct, while more than 60% of correct names were never returned by Soundex. (A.J. Lait and B. Randell, 1996)

Click here to know more about problems & issues with SOUNDEX.

idMatch Systems's Solution: idMatch™

idMatch Systems has created an intelligent software technology for automatically processing multicultural names from different cultures across the globe. We have over a period of time collected huge amount of information. This information backed by our expertise in software development, has led to outcome of idMatch™. We use statistical analysis models and quality data search analytics to improve our product with time. The statistics are used within our product to:

• Generate variants based on the culture
• Recognize phonetic variation name match in the database
• Determine the probable culture of a name
• Determine the probable gender of a name
• Recognize the valid name matches in the database, considering all cultural and phonetic variations possible

idMatch Systems delivers advanced name and identity recognition technology to solve the business problems of automating multi-cultural personal name matching in the database, which in turns allows businesses to:

• Increase Return On Investment (RoI)
• Reduce the potential risks in the business by identifying the fraudulent/terrorist/criminal customers at the business-relationship starting stages itself
• … and many more

idMatch™ software uses the extensive name and other identity data to learn electronically for better precision in name and identity matching and searching with reduced false positives. idMatch™ uses the knowledge of how multicultural names are used around the world. The knowledge of names is used to create the business rules for idMatch™, for processing in each type of the culture. The business rules that are extracted out of the huge knowledge base that we have, allows idMatch™ to handle the various levels of personal name variations and identity data.

idMatch Systems' idMatch™ comes with the built-in linguistics. The cultural information you need to carry out world-class name and identity searching and matching in your own data is provided for you, ready to be fitted out with your business rules and organizational goals. Matching and searching results are ranked, with a score attached, for efficient and automated processing. The system is highly configurable so you get only the names you need.