Research Article
Applying Extensions of Evidence Theory to Detect Frauds in
Financial Infrastructures
Luigi Coppolino, Salvatore D’Antonio, Valerio Formicola, and Luigi Romano
Department of Engineering, University of Naples “Parthenope”, Centro Direzionale, Isola C4, 80143 Naples, Italy
CorrespondenceshouldbeaddressedtoValerioFormicola;valerio.formicola@uniparthenope.it
Received15May2015;Revised23September2015;Accepted30September2015
AcademicEditor:NarayanaswamyBalakrishnan
Copyright © 2015 Luigi Coppolino et al. hisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
he Dempster-Shafer (DS) theory of evidence has signiicant weaknesses when dealing with conlicting information sources, as
demonstratedbypreeminentmathematicians. hisproblemmayinvalidateitsefectivenesswhenitisusedtoimplementdecision-
making tools that monitor a great number of parameters and metrics. Indeed, in this case, very diferent estimations are likely to
happen and can produce unfair and biased results. In order to solve these laws, a number of amendments and extensions of the
initial DS model have been proposed in literature. In this work, we present a Fraud Detection System that classiies transactions
in a Mobile Money Transfer infrastructure by using the data fusion algorithms derived from these new models. We tested it in a
simulated environment that closely mimics a real Mobile Money Transfer infrastructure and its actors. Results show substantial
improvementsoftheperformanceintermsoftruepositiveandfalsepositiverateswithrespecttotheclassicalDStheory.
1. Introduction
Nowadays, mobile communication networks represent a
key enabling infrastructure for inancial service provision
since they ofer signiicant opportunities for increasing the
eiciency and pervasiveness of such services by expand-
ing access and lowering transaction costs. Mobile inancial
services are currently applied to several banking products,
such as deposit and transact products, over-the-counter bill
payments, saving products, intracountry remittances, and
international remittances. In this paper, we focus on Mobile
Money Transfer (MMT) services which allow using virtual
money in order to carry out payments, money transfers, and
transactions through mobile devices.
Suchservicesareanincreasinglyimportantandcommon
payment means in many markets due to the pervasive use
of mobile phones, the steady growth in remittances, and the
needforanelectronicperson-to-personpaymentthatmaybe
analternativeandreliableoptiontopaper-basedmechanisms
likecashandchecks[1].
he growing coverage of cellular networks as well as the
increasing availability of mobile communication services is
enabling the widespread adoption of mobile-based inancial
servicesespeciallyindevelopingcountries,likeKenya,India,
Uganda, and the Philippines, thus creating the opportunity
for a signiicant proliferation of mobile commerce services
as well as for an expanding inancial inclusion. It is expected
thatinthosecountriesmostmobile inancialtransactionswill
concern MMT services in the near term since “unbanked”
people, that is, people who do not have their own bank
accounts, will be attracted by inancial services allowing for
performing payments and remittances by simply using a
mobile phone. he same phenomenon is being observed in
developed countries where citizens are becoming unbanked
due to the widespread economic crisis, and inancial service
providers are beginning to investigate the potential for
adopting these newer payment systems that are emerging in
less developed countries in order to meet inancial needs of
customers.
However, as the mobile inancial services market grows,
the risks related to the use of mobile phone-based payments
increase, since MMT services become the target of more
skilled and motivated attackers, and the amazing volume
of data impairs the ability of the control mechanisms to
timely spot frauds. In one case, a supplier of Mobile Money
services lost $3.5 million due to a single type of fraud.
Like any other money transfer service, an MMT service
is vulnerable to a number of misuses, including money
Hindawi Publishing Corporation
International Journal of Distributed Sensor Networks
Volume 2015, Article ID 980629, 16 pages
http://dx.doi.org/10.1155/2015/980629