LP Eye... For UK Retail Loss Prevention Professionals
Home
LP Vacancies
LP Statistics
Shoplifting
Internal Dishonesty
Burglary
Tiger Kidnap
ORC
CCTV
Intruder Alarms
EAS and Tagging
Product Protection
Uniformed Security Officers
Store Detectives
Test Purchasing
SIA and Licencing
Powers of Arrest
Personal Safety Devices
EPOS Data Mining
Satellite Tracking
Anti Burglary Equipment
LP Training
LP Qualifications
LP Recruiters
LP News
LP Books
LP Resources
Police Force Contacts
LP Forum
Networking

EPOS Data Mining

In most retailers, more than a third of their losses are caused by Internal Dishonesty. Surveys indicate that more than 80% of employee fraud occurs at the till. Most methods enable the employee stealing large volumes of cash whilst still leaving the till balancing. This has to involve elements of bogus transaction detail, which is fraud. Other high value staff dishonesty may involve goods being passed to "customers" without being registered at the till. (This often involves usage of the void function.)

By Data Mining the EPOS Data, a skilled Fraud Investigations Team is able to identify those people committing large scale fraud. the sums of money involved often grow to upwards of 20,000 per year in many cases.
Case Study :-

An employee had devised a method of fraud whereby if a customer purchased high value transactions with cash, he would process and subtotal the transaction, and tell the customer the amount due. After the customer handed over the cash, the employee discreetly hit the "transaction void" button immediately followed by nosale. He placed the cash in the till, and gave the customer the correct change. He kept a tally in his head of how much he'd made the till "over" by. Before the end of his shift, he would remove this amount of cash from this till, and place in in his sock. Typically this was in excess of 1000 per day.

The implemention of data mining led to his arrest. Data Mining typically runs many queries to detect the finger print of fraud within the data. In this case the data mining query was as simple as "show those employees who have completed volumes of transactions voids immediately followed by no sale."
SPSS Clementine SPSS PASW Modeler (Formerly Clementine) is an industry-leading data mining solution that helps fraud analysts and investigators understand people's past behavior and therefore predicts future behavior. Using Clementine, users can access data from various sources across your organization to quickly and easily produce, evaluate, and deploy analytical models, making data mining a cost-effective analytical approach.
PCMS Smart Store PCMS SmartStore - Query Designer for Investigators empowers the Loss Prevention business leaders with the tools to be a specialist support function to retail management teams in order to reduce loss at the PoS Front End. Vision SmartStore automatically identifies suspect transactions most likely to be fraudulent in realtime and alerts users to take action .
IntelliQ Data Mining IntelliQ have provided retail fraud data mining software to many large UK retailers. IntelliQ also provide Fraud prevention and detection data mining systems to many european retailers description to the world, giving a non sales pitch view.
Loss Manager from IDM. A Retail Fraud Data Mining Tool LossManager from IDM a retail-intelligent solution for detecting, analysing and managing internal losses. Developed by experts in retail profit protection, LossManager delivers best-practice loss prevention, train of thought analysis and sophisticated data mining simply and directly to the desktop.
Syarepublic Data Mining SysRepublic - Point of Sale information is supplied to Secure 2007 in Batch or Real Time, alongside other data such as Employee files and CCTV footage. Detecting fraud can be a pain staking and highly technical activity. Secure 2007 simplifies this and allows centralised loss prevention management teams to seamlessly coordinate field officers,to get the most from existing resources. Loss Prevention management and field investigation personnel interact with Secure 2007 through an easy to use web application.