Real Time Querying (PrimeMatch^®)

Why identity searching and matching are important?
A variation in name undetected by the screening application used by government or business causes a terrorist to escape, banker to finance the same delinquent entity, government agency issuing a duplicate passport to a criminal that results in fraud at several businesses. Needless to say Identity searching and matching plays a vital role in today's business for various purposes such as fraud detection, online enquiry, customer search, list-matching, de-duplication and business intelligence applications.

Are there no globally unique identifiers for each entity?
Most common misperception is that there exists a global unique identifier at country level or at company level that will miraculously protect businesses from fraudulent entries protecting data integrity. Not all systems throughout the supply-chain use the same globally unique identification. Most systems do not have very effective fraudulent detection solutions. Some companies implement a unique identifier in one department/geography that is different from unique identifiers at other department/geography. Similarly, companies that have different unique identification systems pose problems at the time of merging or in supply chain that can result in repeated marketing activities to the same customers wasting time, resource and capital. Currently the tools offered by many database vendors not entirely efficient for duplicate record detection.

Even if there are unique identifiers, fictitious accounts are typically created using some other identities defeating the mechanism that relies only on identifiers. Similarly, unless there are joint accounts, banks may not be able to link accounts opened by spouses.

These limitations demand alternative solutions of using name matching or context sensitive matching that can leverage any unique identifier if one is implemented.

What are the alternatives to unique identifiers?
Even though Social Security Number or Passport is available, lot of businesses still rely on matching other key attributes such as name, address, telephone and date of birth. The attributes depend on the type of business and constitute rules. These rules can identify matches even if there are no unique identifiers.

Whether the purpose is an online enquiry or customer search or a batch-matching process of preparation of hit list or de-duplication, name based searches are becoming very prominent. The name of a customer has come to be a critical business asset. As more and more organizations are implementing customer focused systems, quickly and correctly identifying customers enables the organization to gain competitive advantage. Name matching is the task of recognizing whether two names written differently denote the same entity and is widely used in de duplication of databases, data mining, data matching, allotment of unique numbers like Social Security Number SSN of USA, Permanent Account Number (PAN) of India, credit verification while issuing credit cards, bank accounts etc..

What are the problems with alternate identification attributes?
Whether it is names, addresses or other attributes, identity data suffers from either unavoidable error and variation or deliberate attempt to hide as they are recorded in a real world situation. There is lack of support for name matching technology in the native databases and therefore specialist companies provide the solution.

• Example of identity: Shah Rukh Khan
• Order of writing: Khan Shah Rukh
• Splitting/Merging of Parts: Khan Sharukh
• Use of Abbreviations: S R Khan
• Variations in spelling: Khaan Sha Rooq
• Cultural Influences: Sah Ruk Kan
• Missing/Extra Parts: Shah Ruk
• Typos: Shah Rukj Khan
• Noise: Shah RoOk Khan
• These errors in conjunction could make things miserable. Sharo0kh K, Mr.Shark Khan

The problem of name matching appears trivial yet is very complex. Complexity involved in name matching due to variations can appear in the databases and they cannot be addressed through SQL queries alone. Even if the information in the database is deemed to be perfect, both the name to be searched and information available are bound to be different. There are no well defined standards for writing names and addresses.

What are the changes that affect identity search with existing solutions?

Two main driving factors that are game changing in identity search and match technologies are:

• Explosive volume of customer records call for a very cost effective solution that can scale to voluminous data search.
• Flexibility leveraging existing hardware and software infrastructures.
• Dependency on real-time responses requires extremely fast and highly accurate results with no or minimal false positives.

Existing solutions from other vendors are unable to meet these dynamic requirements of the driving factors. They rely on expensive additional hardware infrastructures, tightly integrated into proprietary solutions resulting in vendor lock-in, and cannot scale to tens of millions of records.

The Name Matching solutions provided by the other vendors broadly fall under the following categories - Exact name searches, wild card searches, partial word search, bin based search and fuzzy name matching. These solutions are based on either Phonetic code or similarity metric algorithms that have several limitations. They cannot provide real-time response and give poor accuracy with high false positives resulting in high revalidation effort for the end-user wasting productivity time of the company. Instead of focusing on business, companies spend time on fine tuning and worrying about performance of the system.

Why PrimeMatch^® ?

Posidex Technologies have developed PrimeMatch^®, precisely the kind of technology that helps organizations.
PrimeMatch^® features are as follows:

• Indigenous research based PRIMEMATCH^® is based on number theory that uses radically different approach for searching text.
• The names are converted into numbers using proprietary equations which resolve the match between two names and in the process handles both phonetic and spelling variations.
• Conducts search despite the error, variation, and duplication delivering the highest possible reliability when searching, matching, screening, or grouping data based on names, addresses, descriptions, date of birth and other parameters of identification. It looks for variations of all the parameters thereby virtually leaving any chances of missing any record.
• Keeps the false positives to the barest minimum through a series of steps for organization of data, carrying out analysis based on content of each and every record and a built in mechanism to explore a number of built-in and customizable rules of search.
• Uses revolutionary approach of employing number theory for querying resulting in unlimited scalability, unmatched speed and accuracy over any existing similar product not only in India but perhaps in the world.
• Does not require additional hardware infrastructure reducing CAPEX additionally saving power consumption and data center space.

What are benefits of PrimeMatch^® ?

Posidex's innovative PrimeMatch^® identity search engine precisely addresses the business demands by overcoming limitations of existing solutions from competitors. It is the underlying technology for Posidex Data Management Suite(PDMS) and solutions.

For demo, product and solution evaluations and pricing details please contact us.