Computational Intelligence - February 2017 - 21

B. Doddington's Metric

Traveler impersonation is a key problem of border security
since ancient times. Shifting duties to the machine will not
relieve the border crossing passage of this problem, but impersonation phenomenon should be studied at the level of
Human-Machine (H-M) and Machine-Machine (M-M) interactions instead of the classic scheme based on Human-Human
(H-H) interactions (traveler-border officer). Table 2 provides
some clarification details via comparing impersonation phenomena in H-H, H-M, and M-M interactions (a complete taxonomy is given in [42]). The focus of our study is the inherent
impersonation I-I phenomenon, or zero-effort impostor
attempt known as Doddington's categorization; it is well studied for various modalities [21], [47], including the database
which we used in experiments [57], as well as an extension of
Doddington's categorization [22]. Doddington's phenomenon
is not stable and depends on many factors. Doddington detector utilizes two classes of match scores: genuine and impostor
match scores.
Consider a watchlist operating in an open-set identification
mode when the traveler may or may not be in the watchlist. We
distinguish the following three types of Doddington detectors.
Type-I is based on the evaluation of the deviation between
the score distributions. Each class is evaluated based on the difference between a traveler's score and the mean score of everyone. A traveler is a 'Goat' when his genuine score is relatively
low when compared to the mean genuine score. A traveler is a
'Wolf/Lamb' when his imposter score is relatively high when
compared to the mean imposter score. The Type-I detector
complies to the definition of Doddington et al. [21] when the
score data is considered to be a normal distribution (Figure 2,
first row).
Type-II is based on the analysis of the order of the scores.
Each class is determined by the ranks of the genuine and
imposter scores. Traveler's with low genuine ranks are 'Goats'
and traveler's with high imposter ranks are 'Wolves/Lambs.' The
Type-II detector agrees with the definition given by Ross et al.
[58] (Figure 2, second row).

Traveler
IN
STOP
Authentication
Station

Updating from Physical Sources
Physical Evidence
Watchlist

Virtual Evidence
Updating from Virtual Sources
A1

A2
A3
A4 Storing
Gathering, Processing
and Sharing

Traveler
OUT

Risk
Assessment

figure 1 the model of watchlist technology for e-border applications.

Type-III is based on the examination of the error rates
depending on decision/threshold values, such as 0.5, 0.4, and
0.3. Each class is computed based on the false acceptance rate
(FAR) and false rejection rate (FRR). 'Goats' are travelers that
cause the FRR, 'Lambs' and 'Wolves' are the travelers that cause
the FAR [23] (Figure 2, third row).
C. Unstable Nature of Doddington's Phenomenon
Vs. Dynamic Watchlist Data Structure

Doddington's categorization of biometric traits is an unstable
phenomenon; it depends on various factors such as the quality
of biometric traits, algorithmic features of the processing, and
the decision making strategy at all levels of the processing hierarchy. This is well understood and documented, in particular, in
[22], [42], [47], [57], [59]. Hence, it can be predicted that the
detectors in Figure 2 provide different results due to the above
reasons as well as the different principles of detection.
Various performance metrics are used for evaluating and
choosing the detector. Conceptually, there are two key metrics
for the watchlist check: one for genuine transactions (how well
a traveler matches against their own enrollments) and another
one for impostor transactions (how well the traveler can be distinguished from other individuals). Low impostor match scores
indicate that a user is unlikely to be mistaken for others.

Table 2 Basic impersonation strategies in watchlist technologies [42].
Phenomenon

h-h inTeraCTions

h-m anD m-m inTeraCTions

inherent impersonation as
an error (i-i)

incorrectly recognize
a person, take somebody for somebody
else

inherent impersonation or doddington's metric; the particular
case of an attack Without an impostor is knoWn as a zero-effort
impostor attempt [21]. countermeasures include anti-impersonation, Which includes user-specific thresholding [47], score normalization, and multimodal biometrics [49], [50].

impersonation as an attack
on biometric technology
(i-ii)

pretence to be someone that you are not

examples are plastic surgery [17], color lenses for changing
iris features [19], makeup [20], synthetic or fake fingerprints,
and synthetic irises. countermeasures include anti-spoofing
using texture-based, motion-based, and liveness-based techniques
[51]-[53].

impersonation as an attack
on a risk assessment technology (i-iii)

no analogs

social media mining impersonation is exhibited by distinguishable
patterns such as behavior and activities [35], [54]. an attacker can
impersonate a credible source of information Which is used for
traveler risk assessment. countermeasures partially address
recommender system design [55], [56].

February 2017 | Ieee ComputatIonal IntellIgenCe magazIne

21



Table of Contents for the Digital Edition of Computational Intelligence - February 2017

Computational Intelligence - February 2017 - Cover1
Computational Intelligence - February 2017 - Cover2
Computational Intelligence - February 2017 - 1
Computational Intelligence - February 2017 - 2
Computational Intelligence - February 2017 - 3
Computational Intelligence - February 2017 - 4
Computational Intelligence - February 2017 - 5
Computational Intelligence - February 2017 - 6
Computational Intelligence - February 2017 - 7
Computational Intelligence - February 2017 - 8
Computational Intelligence - February 2017 - 9
Computational Intelligence - February 2017 - 10
Computational Intelligence - February 2017 - 11
Computational Intelligence - February 2017 - 12
Computational Intelligence - February 2017 - 13
Computational Intelligence - February 2017 - 14
Computational Intelligence - February 2017 - 15
Computational Intelligence - February 2017 - 16
Computational Intelligence - February 2017 - 17
Computational Intelligence - February 2017 - 18
Computational Intelligence - February 2017 - 19
Computational Intelligence - February 2017 - 20
Computational Intelligence - February 2017 - 21
Computational Intelligence - February 2017 - 22
Computational Intelligence - February 2017 - 23
Computational Intelligence - February 2017 - 24
Computational Intelligence - February 2017 - 25
Computational Intelligence - February 2017 - 26
Computational Intelligence - February 2017 - 27
Computational Intelligence - February 2017 - 28
Computational Intelligence - February 2017 - 29
Computational Intelligence - February 2017 - 30
Computational Intelligence - February 2017 - 31
Computational Intelligence - February 2017 - 32
Computational Intelligence - February 2017 - 33
Computational Intelligence - February 2017 - 34
Computational Intelligence - February 2017 - 35
Computational Intelligence - February 2017 - 36
Computational Intelligence - February 2017 - 37
Computational Intelligence - February 2017 - 38
Computational Intelligence - February 2017 - 39
Computational Intelligence - February 2017 - 40
Computational Intelligence - February 2017 - 41
Computational Intelligence - February 2017 - 42
Computational Intelligence - February 2017 - 43
Computational Intelligence - February 2017 - 44
Computational Intelligence - February 2017 - 45
Computational Intelligence - February 2017 - 46
Computational Intelligence - February 2017 - 47
Computational Intelligence - February 2017 - 48
Computational Intelligence - February 2017 - 49
Computational Intelligence - February 2017 - 50
Computational Intelligence - February 2017 - 51
Computational Intelligence - February 2017 - 52
Computational Intelligence - February 2017 - 53
Computational Intelligence - February 2017 - 54
Computational Intelligence - February 2017 - 55
Computational Intelligence - February 2017 - 56
Computational Intelligence - February 2017 - Cover3
Computational Intelligence - February 2017 - Cover4
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202311
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202308
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202305
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202302
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202211
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202208
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202205
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202202
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202111
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202108
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202105
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202102
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202011
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202008
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202005
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_202002
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201911
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201908
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201905
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201902
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201811
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201808
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201805
https://www.nxtbook.com/nxtbooks/ieee/computationalintelligence_201802
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter17
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall17
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_summer17
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_spring17
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter16
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall16
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_summer16
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_spring16
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter15
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall15
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_summer15
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_spring15
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter14
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall14
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_summer14
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_spring14
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter13
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall13
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_summer13
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_spring13
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_winter12
https://www.nxtbook.com/nxtbooks/ieee/computational_intelligence_fall12
https://www.nxtbookmedia.com