Mobile device fraud detection using locally stored behavioral information of others

Abstract

A system; method and a device for offline authentication of transactions using mobile device, based on, analytic engine such as behavioral pattern detection are provided. The behavioral pattern can be for a specific person, for group of people with similar characteristics, or a combination of the two. The invention has the advantage over the prior art controlized authenication and fraud detection systems in that it is more precise in identifying and preventing fraud in real time. The precision is better for both customer and merchant frauds. The present invention also requires fewer investments in infrastructure and uses less communication traffic when compared in the prior art.

Description

FIELD OF THE INVENTION[0001]The present invention is directed to the use of mobile devices in offline transactions, and, more particularly, to a system, method and device for self-authentication of transactions.DESCRIPTION OF THE RELATED ART[0002]Mobile devices nowadays are in widespread use. The mobile devices of today have many uses other than plain conversation and messaging. One of the emerging fields of use for the mobile devices is for transactions such as purchasing an item.[0003]The improved ability of mobile devices provides enhanced web capabilities (e.g. internet) and applications. The user interface has improved and thus created a platform for applications, innovative initiatives and new opportunities.[0004]In the context of financial transactions, the mobile device can he used in many processes, when payments by which (in-payments) are one category:of financial transactions implemented by the mobile platform, along with financial services (nit-services) and trade (m-com...

AI Technical Summary

Benefits of technology

This approach enhances fraud detection accuracy for both customers and merchants, reduces false positives and negatives, and allows for more stringent security measures without overloading financial institutions, while minimizing data transmission and processing time.

Problems solved by technology

The SMS communications protocol is inexpensive and relatively simple to use and is now the more accepted method of payment using mobile device, however, the user experience is not adequate.

This technology has limited user interface and user experience is not optimal.

Payment card fraud occurs when an element (e.g. person) creates financial or material gain by the use of payment means or payment means information to complete a transaction that is not approved by the legal account holder.

However, since no statistical model is foolproof, in practice there is always a need to balance between two extremes: a model that will find almost every fraudulent transaction but with many errors (false positive) and a model that will not have many errors but will also miss on real fraud (false negative).

In the current systems it is impossible to check every suspected transaction because it will create an enormous load on the resources of the computing system.

As a result, current systems compromise and do not check every transaction, even though some transactions can be fraudulent.

Ongoing struggle with fraud drove their extent down.

This is because the investment in infrastructure in order to verify small amount transactions would not be cost effective compared to the gain.

During the check, the server has to retrieve all the needed information needed for processing and perform a large amount of complex mathematical calculations.

In practice, these servers are very expensive.

Therefore the issuer compromises on the quality of the statistical models and the quantity of the cheeks.

The result is that the level of coverage and accuracy are insufficient and there are many mistakes:

In practice, due to the low level of accuracy, transactions are rarely blocked.

However, when self-authentication (offline approval) process estimates a possibility of fraud usage, online approval for more detailed statistical analysis processing is requested from a remote computer.

None of the current technologies and prior art, taken alone or in combination, does not address the issue of offline authorization, self-authentication and fraud detection of a transaction, e.g. there is no handling of the security aspects of the transaction without requesting the bank or the credit card company for approval.

There is also no solution to the issue of using a statistical model with lower suspicious rate then current statistical models without blocking the transaction or the card.

Images