Hiding Behind a Name…

Umar Farouk Abdulmutallab (Arabic: عمر فاروق عبد المطلب ); also known as Umar Abdul Mutallab and Omar Farooq al-Nigeri is popularly referred to as the “Underwear Bomber”.  He is a Nigerian man who, at the age of 23, confessed to – and was subsequently convicted of – attempting to detonate plastic explosives hidden in his underwear while on board Northwest Airlines Flight 253, en route from Amsterdam to Detroit.

Intentional deception or even unintentional errors in the recording of a name can confound simple deterministic matching systes. A simple transposition or even a legitimate variant of a name will not be considered when searching for a record in most existing Record Management Systems (RMS). To overcome this, significant research and development into the creation of “fuzzy” name matching algorithms have attempted to overcome the challenges in name matching.  The need for more sophisticated, flexible searching methods has become more pressing as our population diversity increases, information is shared across jurisdictions, and individuals provide incorrect or misleading information at the time of booking.

IMT Intel:ID™, a purpose-built entity resolution and non-obvious relationship solution for law enforcement, leverages a process that connects the dots between new data and existing data.  IMT Intel:ID forms contextual relationships (known as Context Accumulation) to resolve duplicate person records and establishes links between those individuals to uncover non-obvious relationships. The social identity of a person is formed by the interactions they have with other individuals.  This “social context” holds valuable information, which is difficult to fabricate or alter, because the social network reinforces the identity of the person analyzed. This offers actionable intelligence and allows investigators or front-line officers to gain insights and increase the likelihood of identifying persons of interest and the individuals they may know.

The Context Accumulation process is greatly enhanced by leveraging the power of IBM® InfoSphere® Global Name Management (GNM).  GNM is an advanced technology, developed over 20 years of real-world collaboration with linguists from around the world, for improving name matching across diverse cultures that leverages the latest advances in computational linguistics. This approach is based on the application of statistics, mathematics, linguistics research and computational expertise to enhance name matching. In the example of Umar Farouk Abdulmutallab, IBM’s GNM found almost three thousand legitimate permutations of the recorded name based on acceptable variants of those name tokens.  This gives new meaning to hiding behind a name.

IMT Intel:ID seamlessly integrates with GNM to search for names in a database or list. Based on empirically-derived linguistic rules, mapped to the specific culture of the searched name,  GNM provides the best possible matching, searching, parsing, and scoring. This capability takes into account the most likely cultural-based variations. In turn, GNM enhances the accuracy of name searching to improve law enforcement identity verification initiatives. This helps address the problem of differences due to transliteration, as well as the profusion of naming and syntactical schemes that can make it difficult to distinguish suitable candidates for matching purposes.

Intel:ID leverages the following key GNM libraries to strengthen its resolution capabilities:

  • NameClassifier™ – Identify and classify the most likely culture (ethnic category) of a name, including the countries in which the given name or surname is most often found
  • NameGenderizer® – Recognize and report the relative frequencies of gender (male or female) associated with given names
  • NameParser® – Parse personal names into surname and given name components
  • NameVariationGenerator® – Generate lists that contain variant forms of the components (given name and surname) of a name
  • Name Matching capabilities – Match names regardless of spelling, typing errors, cultural variations, and structure. Match names on both pronunciation and orthography, with the closest matches returned first

The Enhanced Name Capability (ENC), which is an add-on to Intel:ID, adds additional GNM functionality that boosts the accuracy of any search that includes a name as part of the search criteria.  ENC integrates GNM’s NameHunter® seamlessly into the Intel:ID environment to allow investigators to cast a wider net – yielding better result sets – which enhances the accuracy of name searching and the overall quality of identity verification initiatives. Using the above noted components to search a list of names via culture-specific search strategies Intel: ID continues to leverage its robust non-name attribute matching capabilities.  This enhancement should be considered when most of the searches within Intel:ID include name as a constraint, and especially if it is often the only constraint.

When a person of interest can be tracked and linked with more precision and speed than current tools are capable of, law enforcement can proceed with more confidence – establishing a higher level of security and safety for all citizens while increasing officer safety.

To learn more about IMT Intel:ID™, please contact Mark Holmes at: mark@imt.ca