Fractals Digital Fraud Brochure

Fractals Digital Fraud Brochure
For more information visit or contact NCR today.
Executive Summary
Fraud solutions are seeking to become more
enterprise-focused and channel-agnostic.
Fraud patterns are evolving and this trend is
supported by the adoption of advancements in
technology and financial services offered
through Online and Mobile channels. The
industry has reached a Digital Age, with risks
and exposure behind every interaction in the
Internet of Things (IOT), which drives increased
points of data transmission and vulnerability.
Figures from Gartner research forecast that, by
the end 2016, there will be around 6.4 billion
connected ‘Things’ in use around the world, a
30 percent increase over the previous year.
However, by 2020, this figure is expected to
rise to almost 21 billion.
As consumers and businesses look to increase
activity through digital channels, fraudsters
are leveraging weaknesses in these channels,
to gain access to customer data and financial
services for money movement. It is often
difficult to know if a particular transaction or
event is originating from a trusted customer or
a sophisticated fraudster; the latter may hide
their true location, using compromised login
credentials, stolen devices or stolen account
data. Fractals Digital Fraud Service
encompasses both non-financial and
financial information, to create a holistic
view of customer activity, enabling fraud
strategies to identify suspicious activity
as proactively as possible.
Fractals Digital Fraud
SaaS Solution
The Fractals Digital Fraud Service aims to
provide Financial Institutions with front-end
detection capabilities for their online and
mobile channels. Fractals Digital Fraud
Service consumes and analyzes related session
and payment origination data to execute
an informed decision on the risk of every
event. This means analyzing every event for its
inherent risk, based on a number of factors,
including characteristics of the event – e.g. IP
address, amount – customer specific profiles –
e.g. typical IP address for the customer – and
analysis of the recent history of the customer.
This approach allows a financial institution
the ability to look for risk, or lack of risk,
and make a decision about the level of
authentication required for online or mobile
customer interactions. This promotes a
strategy which mitigates the risk to an
acceptable level, while inconveniencing the
customer to a minimum – all in real time.
“The Fractals Digital Fraud
Service aims to provide Financial
Institutions with front-end
detection capabilities for their
online and mobile channels.”
As a SaaS solution, the Fractals Digital Fraud Service provides turn-key features
and functions for Financial Institutions that address the following areas:
User Administration and Access Provisioning with Multi-Organization Support
Rules-based Detection combined with Machine-Learning Analytics
Alert Management Workflow
Dashboards & Operational Reporting
Solution Training & Product Support
Fractals offers distinctive combination of user-configurable rules and self-learning
capabilities. At its core, the Fractal Digital Fraud Service combines the power of
intelligent, machine-learning analytics, highly configurable profiling, and a rules
engine to put users in charge of their fraud prevention and detection operations.
The Fractals Digital Fraud Service enables real-time, in-flight fraud blocking.
For example, “Login” messages can be routed to Fractals in real-time, for risk
scoring and rule detection. Based on the risk score and outcome of triggered rules,
Fractals reply with a message that recommends challenging the user at login.
This approach provides organizations the ability to prevent fraudulent logins
through a stepped-up authentication, in real-time!
Fractals also supports near real-time and batch detection modes for processing
and alert triage. The diagram below illustrates a high-level view of core capabilities,
with key elements of Fractals fraud detection highlighted in yellow and described below:
Fractals Fraud Detection and Prevention Solution
Profiles, Rules
and Models
External Datasets
Transaction Data Source
(e.g. Authorization System)
Fraud Score/
Data Input
via Fraud
Hub (FIH)
Fraud Score/
Learning and
the Adaptive
Digital Fraud and User
Controlled Rules
Rules form the core of the power and
flexibility of Fractals Digital Fraud Service.
Out of the box rules are available, but
Financial Institutions have the ability to
create, test and implement their own fraud
detection strategies. Fractals Digital Fraud
Service gives the user the ability to create
rules of any sophistication, easily, quickly
and with very little training. Users can easily
add and modify rules, test them, check for
similarity with other pre-existing rules and
set rules to production status, all within a
matter of minutes.
Advanced Dynamic Rules
Rules can make use of data within a single
transaction or within a run of transactions
and can use functions and statistics which are
derived from data in the system. Examples of
powerful functions include:
Velocities, such as number of transactions or events in the last hour on an account or
Aggregate value of transactions in the
last day
These functions can be compared with other
functions such as a long term (e.g. 90-day)
moving average of daily spend on an account
or user profile changes over a 30-day period.
The use of these powerful functions to perform behavioral analytics enables the engine
to detect deviations from the historic customer
profile – for example to assess whether the
spend in the last hour is greater than 3 times
the 90-day moving average for daily spend. If
this scenario is breached, the Fractals Digital
Fraud Service can generate an alert and the
transaction or event in question cold be held
or declined.
Advanced Models and
Self-Learning Strategies
Dashboards and
Alert Management
The Adaptive Classification Engine (ACE)
model is a sophisticated Bayesian model
with self-learning capabilities. It provides
significant benefits to an organization and
major advantages over previous generation
neural network technology or standalone
rules-based detection. The model is used
to analyze every integrated event
message through thousands of variables
and produces a score based on the
likelihood of fraud.
A user-defined dashboard can be configured
and built to display aggregated alert-related
data for quick reference. Dashboard graphs
can be configured by the user and allow
information to be separated by current day
operational view or historical data view (e.g.
past week, past month). These graphs can be
built and configured based on the Fractals
data model and updated to meet customer-specific requirements.
That score can also be used in the rules
engine to enhance rules and reduce
false positives.
“(ACE)…provides significant
benefits to an organization
and major advantages over
previous generation neural
network technology…”
User-friendly screens help end users manage
their workflow, analysis and review of alerts,
control lists such as blacklists or whitelists as
well as initiate and track any customer contact
via the Alert review screens. All our Graphical
User Interfaces (GUIs) are built on the newest
technology and are able to run on most
commonly used browsers.
In Summary
• Fractals is proven to detect and prevent fraud – it is tried, tested
and live at high profile customer reference sites.
• Fractals offers cross-channel capabilities that provide economies
of scale in operation, infrastructure and training.
• Fractals is easy to use and maintain.
• Fractals is already successfully deployed as a multi-institution, multi-currency combined issuer and acquirer system in several sites.
• Fractals can be delivered and put into production rapidly to
provide immediate improvements in detection performance
and a fast Return on Investment (ROI).
• Fractals user interfaces enable you to have complete control over your system without the need for technical expertise.
• Day-to-day operation of Fractals requires no vendor involvement and no IT staff involvement.
• Fractals operates in true real time, in-flight, transaction blocking mode: essential for maximizing fraud savings.
• Fractals self-learning, intelligent models, deliver class leading
detection performance with extremely high rates of detection
with low customer impact and limited model tuning required.
• Fractals leverages the Fraud Integration Hub (FIH) to integrate
rich third party data for use in rule detection and model analysis (e.g. Geo Location for IP Address).
Fractals Alert Management workflow is simple and easy to use.
Alert and transaction details are displayed in a meaningful way
for analysts to quickly make decisions on suspect transactions
without the need to reference multiple source systems.
Operational Reporting and Dashboards are configured to meet
business needs. The ability to quickly see aggregated and
graphical representations of data can provide fraud supervisors the ability to execute trend analysis and optimize staffing levels.
Why NCR?
NCR Corporation (NYSE: NCR) is the global leader in consumer transaction technologies, turning everyday
interactions with businesses into exceptional experiences. With its software, hardware, and portfolio of
services, NCR enables more than 550 million transactions daily across retail, financial, travel, hospitality,
telecom and technology, and small business. NCR solutions run the everyday transactions that make your
life easier.
NCR is headquartered in Duluth, Georgia with over 30,000 employees and does business in 180 countries.
NCR is a trademark of NCR Corporation in the United States and other countries. The company encourages
investors to visit its web site which is updated regularly with financial and other important information
about NCR.
NCR Corporation
3097 Satellite Boulevard
Duluth, Georgia 30096
©2016 NCR Corporation
Patents Pending
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF