Anti-Fraud Algorithm



WR SWAT

Basic
Joined
25.12.19
Messages
15
Reaction score
0
Points
1
Fraud Screening

Our transaction goes through "pipeline". It consists of the following steps:
Rules System

Data analysis of card and cardholders via using global fraud screening service (Ethoca.com)
Multivariate Regression investigation of in-house database consisting records of transactions

Here is a detailed version of the above stated subsystems:

Rules system:

Fraud rules logic executed in stored procedures via Oracle DBMS. This facilitates addition and modification of rules sans service downtime. We focus on retrieving additional information from the MaxMind credit card fraud prevention service prior to passing order via rules chain. MaxMind works towards returning the following data to the gateway:

Cardholder located in high-risk country. The following nations are recognized as high risk:
Egypt
Ghana
Indonesia
Lebanon
Macedonia
Morocco
Nigeria
Pakistan
Romania
Serbia
Montenegro
Ukraine
Vietnam


Whether nation of IP address matches billing address country (any mismatch is equivalent to higher risk)
Country Code of the IP address
Accurate distance from IP address to Billing Location (km). Large distance is considered as higher risk
IP address’s estimated State/Region
Estimated City of the IP address
Estimated Latitude of the IP address
Estimated Longitude of the IP address
ISP of the IP address
IP address Organization
IP address is behind an anonymous proxy(anonymous proxy is regarded as very high risk)
Probability of IP Address being an open proxy(transparent)
E-mail from free e-mail provider
E-mail in high risk e-mails database
UsernameMD5 input in high risk usernames database
PasswordMD5 input in database of high risk passwords
Nation of issuing bank based on BIN number matches billing address country
Bank’s Country Code that issued the credit card based on BIN number
Issuing bank matches entered BIN name
Bank Name that issued the credit card based on BIN number
Customer service phone number matches BIN phone
Customer service phone number listed on back of credit card
Phone number of customer is in the billing zip code
Shipping address is in database of known mail drops
Billing city and state matches ZIP code
Shipping city and state matches ZIP code


As soon as we collect all requisite data, rules in chain apply to order data in sequence, rising or declining of overall fraud score.

Our Basic Rule Chain includes a set of following Rules
Cardholder country rating (global list)
Cardholder country rating set up by (merchant)
Cardholders IP found in black lists
Cardholders IP range found in black list
Cardholders email found in merchants black list
Cardholders email found in global black list
Cardholders email found in forbidden email providers list
Card PAN is not present in global black list
Card PAN is not present in merchants black list
Cardholders address should not appear in global black list
Cardholders addresses should not appear in merchant black list
Amount of order does not exceed global purchase limit
Amount of order amount does not exceed purchase limit of local (merchant)
Single PAN daily turnover does not exceed global daily limit
Single PAN daily turnover does not exceed local (merchant) daily limit
Billing address daily turnover does not exceed global daily limit
Daily turnover of Billing address does not exceed local(merchant) daily limit
PAN number brute force check
Expiry date brute force check
CVV brute force check


Our fraud officer makes successful attempts to keep tabs on all transaction flow and modifies any existing rules. It also implements the latest ones in order to get maximum efficiency in fraud prevention procedure.

Analysis Transaction History (in-house service):

Rules assessment and verification of transaction data against existing transactions helps in obtaining the most accurate results and fraud decisions.

Analysis Transaction History (external service):

In case, the in-house transaction history does not display any signs of fraud, external database enters gets into action. Ethoca360 fraud detection services control Global Fraud Alliance member data. This is also recognized as the world's biggest private database of first-rate e-commerce transaction data. The pooled data resource consists of 100s of times more data as compared to the most Alliance members possess on their own. This works towards enhancing their fraud detection precision by up to 10 times.
 
Top Bottom