Browse Prior Art Database

Detecting fraud using information on account holder collected outside the operation of the account

IP.com Disclosure Number: IPCOM000035207D
Original Publication Date: 2005-Jan-20
Included in the Prior Art Database: 2005-Jan-20
Document File: 2 page(s) / 44K

Publishing Venue

IBM

Abstract

The accuracy of the fraud scoring process can be improved by including 'out-of-band' data from the activity of the account holder such as current mobile phone location in the 'fraud scoring' process, . The approach proposed here uses 'out-of-band' data from the activity of a given transaction (such as an in person credit card purchase or online banking transaction) and add this to the existing fraud scoring mathematical models.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Detecting fraud using information on account holder collected outside the operation of the account

Detecting credit card and other fraud (e.g. online banking) using a number of mathematical models to build a 'fraud scoring' output is well established. One gap in the existing solutions is that the existing fraud scoring models rely on the data collected as part of the transaction only, and do not rely on real or near-real time 'out of band' data on the account holder at the time of the transaction/authentication/scoring.

    For example, when a credit card processor or online bank attempts to use a mathematical model to flag likely fraudulent activity, they may use some of the following information:
1. mathematical models based on account activity (previous purchase amounts, purchase retailers, account address)
2. mathematical models based on data supplied to a retailer (email address, IP address of requesting internet browser, delivery address, purchase amounts, type of goods, length of standing of account)
3. mathematical models based on other activity across multiple retailers (purchasing patterns across multiple retailers)

    By including 'out-of-band' data from the activity of the account holder such as current mobile phone location in the 'fraud scoring' process, the accuracy of the fraud scoring process can be improved.

    The approach proposed here uses 'out-of-band' data from the activity of a given transaction (such as an in person credit card purchase or online banking transaction) and add this to the existing fraud scoring mathematical models for determining a new fraud scoring model.

  This 'out-of-band' data could include, but is not limited to the location of the account holder's mobile telephone using GSM location techniques at time of transaction the last journey on the account holder's electronic public transport pass the last telephone number for which a telephone call of the account holder's mobile has been made to the last time and location of the use of a corporate security keycard or instant messaging system the time and physical location where an account holder has logged onto a roaming telephone system such as a 'hot-desk' environment or a VoIP softphone requiring authentication

The advantage of using such 'out-of-band' data is that: the authentication of an account holder can be inferred from the authentication of the same indivdual inherent in other systems (such as requiring a PIN number...