Facilitating Trade:
Improving Customs Risk Management Systems
In the OIC Member States
41
information from the common domain (i.e., information exchanged with other EU MS CAs). Data
analysis helps the CRM in the detection of deviation by providing a system foundation based on
a combination of indicators, profiling of similar entities (people, means of transport) and
commodities combined with the analytical tools and data mining algorithms. By using the CRM
system, customs specialists can define peer groups and compliance models for declared
consignments and people. They can perform a retrospective comparative analysis of customs
data for a specific type of consignment, based on attributes such as weight, country of origin or
shipping dates - that assigns risk scores using statistical methods. These basic indicators are
entered into the CRM system and used to calculate the score. Use/combine of additional
indicators can trigger a comparative analysis. The various indicators combined with the
historical non-compliance models can assist in the creation of risk profiles. The CRM should be
able to calculate the value for each indicator in the non-compliance model to a standard numeric
score by comparing an individual indicator value against another consignment that is being
profiled. The more a consignment suspiciously deviates from its peers, the higher is the assigned
score. Finally, a hierarchy of scores or “ground for suspicion” is created from the profiles, with
high-risk consignment being flagged for selection and review. Data mining drives the success
directly affecting the “hit rates” of inspections of targeted consignments. Effective targeting,
through the ability to produce accurate and timely decisions about potential fraud violations can
help in improving the regulatory enforcement and resource deployment. At the same time,
compliant cargo and passengers can be quickly processed.
3.1.5.2 Selection of high-risk shipments with data analysis tools
The BI system can assist the CRM analysts to enhance case selection, and proactively prevent
fraud and other regulatory violations. The data analysis tools provide the CRM analysts with
more efficient ways to manage and mining the data to identify importers/exporters that are
misdeclaring their consignments. Exchanging experience within the field of fraud prevention,
the OIC CAs can recognize similarities between CAs and be able to leverage existing, proven
technologies and methods to develop a CRM approach that can be efficiently applied in their
national domains. Many elements would need to be adopted and designed according to the
legislative and regulatory framework for specific CA.
CRM gather a wide variety of structured and unstructured data. The DW/BI is the solution to
manage such a complex data layers. The CAs must have a clear understanding of what drives
their business and technological needs. Examples of the structured and unstructured data can
include:
 Historical crime incidents: location, crime type, severity, victims, suspects, convictions,
criminal behaviors, and attributes;
 Enabling factors: place, route, time of year, month or week;
 Trigger events: holidays, weekends or working day;
 Unstructured data: pictures, audio/video, and text contained in irregularities reports, e-
mail and open source.
This information is critical for analyzing interactions and uncovering the attitudes, desires, and
motivations of entities to get the details ahead with offenses. is presenting the DW
setup (Customs, OGAs, other sources data) for analysis services and predictions as the output of
the data mining.