Fuzzy Matching
conditions to produce a more
thorough outcome. Matching is a
more sophisticated and successful
way of finding and accessing large
amounts of stored data of all kind of
formats: text, structured or even in
image or biometric coding.
How Does Fuzzy Matching
Work?
During a fuzzy matching process, each
of the criteria is scored against the data
in each field in the database. The scores
per case are than weighted, taking its
relative importance into consideration.
More important criteria have a bigger
impact on the total match score. Finally
the resulting match scores are ranked.
The highest (or lowest) scoring data-
records can now be displayed, often only
if they exceed a certain threshold. The
whole matching process (matching one
passenger against the full database) is
completed in under a second.
Fuzzy matching (as opposed to
database searching) is based on a
number of principles:
Intelligent passenger differentiation will go a long way to ensuring that passengers
circle6 Some search criteria are more can, once again, move quickly through checkpoints and enjoy the airport experience.
important than others: different search (Credit: "Views From The Left Seat" blog)
criteria may be weighted
Some search criteria are Some data may not be
circle6 Some data may not be present more important than others accurate
In differentiating passengers, for example, Textual data, such as locations, and
circle6 Some data may not be precise (either the booking date and payment method numbers such as birth dates or
in numbers, text, geographical may be of less importance, while the passport numbers are all very prone to
indication or even imagery) country of origin (airport of departure) may typos. Dates, for example, are written
be an important factor. In the fuzzy in different ways either side of the
circle6 Some data may not be accurate matching system all different matching Atlantic (eg. 12/4 in the United
criteria should be configured with a factor Kingdom is 12th April, whilst it is
Fuzzy matching for passenger in order to obtain a weighted score. December 4th in the US). Regular
differentiation can be conducted in database searching was never
two ways: Some data may not be developed for non-exact searching,
The first way is to match the known data present thereby failing everything which is not
on the passenger against a number of In any Passenger Name Record (PNR) exactly as per the search criteria.
profiles. Such profiles are combinations of a lot of data fields may not be Also, since the goal of passenger
search criteria which are considered to present. This is caused by the sheer differentiation is to identify possible
contain elements of risk. For example: drug number of different parties who are fraud or trespass, the differentiation
traffickers are known to travel certain providing data for PNR’s and the fact processes and system have to be
routes, and may differ from regular that PNR’s in most reservation designed based on the fact that some
travellers in terms of payment method, systems contain either unstructured data maybe inaccurate intentionally.
time between booking and travel date. fields or even data that is used solely
Profile records are created based on expert within a single airline. Some data may not be
knowledge of security staff at a certain Different governments currently precise
airport and can be automatically created require airlines to provide different data Some data, both in the database and
from watchlists or specific interest groups (in different formats). Unfortunately, in the criteria may not be precise. If
of immigration or customs agencies. these requirements are not uniform. We you are looking for a man between 30
The second method is to match for may expect that, in the future, different and 40 years old, would you
similarities of known perpetrators or data-sets are collected at different automatically assume that someone
for dissimilarity with the group of times by different parties. Missing data aged 29 is not worth further
other “travellers” fields are inevitable. investigation? Humans do not select
24 April 2008 Aviationsecurityinternational
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