CasinoCityTimes.com

Gurus
News
Newsletter
Author Home Author Archives Search Articles Subscribe
Stay informed with the
NEW Casino City Times newsletter!
Newsletter Signup
Stay informed with the
NEW Casino City Times newsletter!
Christian Plaichner Archives
More Strategy Experts

Christian Plaichner Gaming Guru

Christian Plaichner
 

Technology: Complex Events Processing

9 July 2008

By Christian Plaichner

With which methods do risk managers try to make headway against the multitude of threats from insider betting, false identities, multi-accounts, retroactive payments, unsecured credit cards, syndications, and money laundering? Up to now, approaches have attempted to assign illegal activities to a particular culprit via clear identification of the customer with solutions like I.P. tracking, geolocation, and hardware fingerprint, in order to stop them.

A second approach is the well-planned setting of various limits in order to limit damages and extreme values. Complex Events Processing, or C.E.P., uses user-behavior analysis for fraud recognition as well as the identification of patterns and correlations -- and this in real time, while transactions occur. In contrast to pure data mining that only analyzes historical processes, here, suspicious activities can be responded to and prevented before damage occurs.

Easy Detection of Fraud Patterns

Technically, C.E.P. software works as an "intelligent bypass" to the existing I.T. systems, such as a transaction platform, bookie software, payment system, customer retention management system or security system for customer identification. Via adaptors and interfaces, relevant data and events are collected and analyzed in real time.

The processing of the data occurs in cycles, each with five steps: sense (receive and unify events); interpret (generate business information); analyze (recognize situations and recognize exceptions); decide (make decisions); and respond (implement decisions). This event-processing model is continuously adjusted based on the updated data foundation, making it possible to detect exception situations such as fraud patterns in time.

Example for Event-Driven Rules for Combating Fraud

Preconfigured templates are adjusted to the specification of each company; the experience of risk managers goes into the simple graphical creation of rules and assumptions for fraudulent activities. These rules are constantly being relaxed or strengthened, depending on their effectiveness and relevance. With this, the fraud-recognition system automatically adjusts itself to the rapidly changing patterns. The rules control the flow of events and lead to automatic decisions and actions, as in this example for the prevention of insider betting.


Click to Enlarge Image

  • When bets are placed, it should check whether they have anything to do with an illegal insider bet

  • If this can be clearly determined, automatic corrective measures are initiated: rejection of the bet, account locking, initiation of manual inquiries, changes in the betting odds

  • At the same time, measured values for various indices or the automatic adjustment of the rules can be set

  • If the bet cannot be clearly attributed to an insider, a “fraud status” is conveyed to the pattern and the attributes of this bet and, afterwards, the indicators of the bet are compared with known fraud patterns

  • If this behavior corresponds to dangerous behavior or a suspicious pattern, responses such as the rejection of the bet, blocking the customer's account, an alarm or the initiation of manual inquiries can be triggered

    Interplay between Online and Offline Activities

    The event processing is not only limited to the online sphere, it can also incorporate data from offline activities into the analyses and be brought into context. This can be done in a casino -- in relation to data from access systems, cash transactions, and profit margins, for example -- in order to recognize unusual and suspicious patterns in real time.

    Business-Critical Knowhow Remains in the Company

    The use of C.E.P. systems for fraud detection in the gambling sector brings significant improvements and benefits for providers. The software is supplied as a "tool" from external partners, and the creation of the rules for fraud detection is carried out using the combination of preconfigured templates with the internal risk managers’ company-specific knowledge. A further adjustment of the rules to changing threat scenarios can be easily carried out by internal employees without any programming knowledge - essential know-how and all data remain protected in the company.

    Up to 20% Fewer Cases of Fraud After Three Months

    Pioneers among international gambling providers have been very successfully relying for some time on the new technology, thereby achieving significant cost savings and guaranteeing their legitimate customers secure business transactions.

    Example of a roject for fraud detection using C.E.P. with an international betting provider:

  • Project duration less than three months

  • Training of internal staff for six man-days

  • In the first three months after the project, the fraud cases that appeared could be reduced by more than 20 percent

  • The effort for manual tracking of fraud cases can be reduced by more than 25 percent

  • The manual effort to amend the betting odds and betting offers can be reduced by nearly 30 percent

    Innovative Visualizations Easily Capture Knowledge

    Currently on the market, standard systems for the configuration of various limits can be massively upgraded and automated through the use of C.E.P. technologies.

    With the visualization in the C.E.P. system, the risk manager receives a clear picture of suspicious patterns and correlations (e.g. sleeper or putter-on patterns) and can quickly and efficiently respond to new threats.

    As an additional benefit, a C.E.P. system can be used for the analysis of customer behavior and the resulting creation of marketing and sales activities, without additional expenses.

    Large benefits are often made possible through the use of well-known technologies for new issues or problems -- with the use of C.E.P. for preventively combating fraud, a winning combination appears to be successful.

    Christian Plaichner
  • Christian Plaichner