Fuzzy Matching
“…12/4 in the
knowledge of criminal trends of that resources industry, it is relatively new
specific airport. Human evaluation will for security applications; employment
United Kingdom
never be replaced by a fully automated agencies use fuzzy matching for
system, but combining human finding resumes for their vacancies.
is 12th April,
observations with a matching against Some government agencies in the
profiles increases the chance that immigration and public safety
whilst it is
somebody will be stopped while crossing industry are piloting or implementing
the border illegally or while carrying illegal matching tools for differentiating
December 4th in
goods. Fuzzy matching may be used at the groups in order to work more
gates as a decision-support tool to assist in efficiently and give the right amount
the US…”
determining whether or not a passenger of attention to appropriate clients.
should be investigated further. There is no standard yet on the
Passenger differentiation may also be criteria to match against and, even
used before the passengers even arrive whether matching against profiles,
at the airport. The screening process, as against similarities or a combination
in that rigid way, but regular database conducted currently at airports, can of both will produce the best results.
queries do. influence the flight’s operation. For Another challenge for passenger
In order to model the searching example, if a passenger is denied differentiation is similar to the
process after human behaviour you will boarding, their luggage has to be challenges that Advanced Passenger
need flexible numeric boundaries. retrieved from the aircraft and may Information (API) brings to airlines
Everything within the boundaries scores cause a delay. Predicting this possibility and governments: how to get the right
100%. For example, if the man is 35 ahead of time could ensure that the data, at the right time and in the right
years old in aforementioned example, baggage of such a passenger be place. Although fuzzy matching allows
the score is 100%, while the score handled differently. This will require a for partial data, the best results are
rapidly decreases on either side of the change in some processes but that’s obtained when the data is timely and
boundaries. Someone aged 45 may exactly what is needed. as complete as possible.
score only 50%.
Some intelligence is needed to relate Trusting the Trusted Conclusion
values to each other. For example there Traveller There will be an increasing need to
isn’t a big difference between a banker A relatively recent trend in the world of optimise the screening process, because
and a financial advisor or between a aviation is the development of frequent of the constraints on physical working
carpenter and a constructor. In fuzzy traveller programmes (not the airline space and resources, let alone increasing
matching this is called “affinity”, and is reward schemes, but the passenger governmental requirements and desired
implemented using matrices. Within facilitation variety in use at immigration passenger flows. An instrument that may
passenger differentiation this may be controls), which help passengers to be used to conduct a smarter screening
implemented for grouping locations, quickly pass through checkpoints. This process is to differentiate passengers
carriers, payment methods, synonyms of can be accomplished because of the into different groups for quick or
first names, the aforementioned fact that the passenger is identified intensive screening.
professions or other criteria that relate using biometrics and some enrolment Fuzzy matching technology seems a
to each other. For example, customs process (including pre- screening) has pre-requisite for passenger
departments may be interested in already taken place. As the groups of differentiation in the transportation
passengers who are animal traders, yet pre-screened travellers becomes bigger sector because of the timeliness,
regular traders or event veterinarians and bigger, these “trusted “groups may accuracy and variable nature of the
are related professions. In fuzzy in themselves become a security risk. available passenger data. Although
matching an affinity matrix will be set up Due to the lower number of mature in other industries, the
between the professions, and whereas confrontations with security staff, the application of fuzzy searching in the
the animal traders would get a 100% status of being a “trusted traveller” is aviation industry is still uncommon,
score, traders and veterinarians may very attractive to those with bad intent. yet the indicators are that its absence
only contribute 60% to the total score. In addition, there may be a false sense will not be for long.
of trust, because of the screening done
Fuzzy Matching Will at enrolment. It would be better to
Support Human Screening evaluate the traveller data each time The author is a senior consultant and
Rather Than Replace It someone passes a biometric project manager working at Accenture
Security personnel involved in border identification gate and using fuzzy within the aviation industry and
management, customs or policing already matching this can be achieved. immigration, justice and public safety
differentiate people on the move. The domain. Accenture is a global
origin of a flight, passport data, age, Matching Challenges management consultant and system
luggage, clothing style and behaviour may Although fuzzy matching technology is integration company with more than
be indicators for a particular passenger mature (it has already around for 175,000 employees worldwide. The author
being subject to questioning or physical more than 10 years) and has been can be contacted via email at
search, especially if this is combined with applied successfully in the human martijn.moret@accenture.com.
Aviationsecurityinternational April 2008 www.asi-mag.com 25
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44