RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

MTS Big Data Center: Antifrod Platform

Product
Developers: Mobile TeleSystems (MTS)
Date of the premiere of the system: 2023/11/24
Last Release Date: 2024/03/22
Branches: Internet services,  Information security
Technology: Information Security - Fraud Detection System (Fraud)

Main article: Fraud Detection System

2024: Launching a solution to combat phone scammers

PJSC MTS"," a digital ecosystem, on March 22, 2024 announced the launch of a solution to combat telephone fraudsters. This anti-fraud platform allows you to to banks receive in real time information about suspicious actions based on almost a hundred different factors. With the help of the MTS service, the accuracy of determining fraudulent actions increased by 45%.

According to the Central Bank of the Russian Federation, in 2023 the volume of funds stolen by fraudsters in Russia increased to 15.8 billion rubles, which is 11.5% more than in 2022. The service developed by the Big Data MTS Scoring team will allow banks to significantly reduce the number of fraudulent cases. First of all, this concerns the registration of loans and the transfer of funds, as well as theft of personal data of customers using social engineering methods.

Anti-fraud algorithms based on machine learning label subscriber numbers for banks that are subject to telephone fraud. Each client is analyzed by the system in two models. The first model compiles a customer profile based on his actions, using dozens of different characteristics. The second model analyzes in real time the flow of events, calls, SMS and other client actions, identifying various kinds of patterns that fraudsters use. The composition of these two models helps to determine whether the client is a victim of social engineering. The platform is currently being tested in several banks.

File:Aquote1.png
Over the past few years, we have made significant progress in the development of anti-fraud solutions. If until recently such systems could take into account only a few factors, such as, for example, replacement numbers, now we are training algorithms already on a hundred different indicators, and their number continues to grow. All this makes it possible for banks to receive the most accurate information. In the future, we plan not only to develop the solution as an external product, but also to implement it in our mobile network to protect subscribers, "said MTS Big Data Director Victor Kantor.
File:Aquote2.png

2023: Launch of an anti-frode solution on marketplaces, message boards and online stores

MTS, digital ecosystem, November 24, 2023 on the launch of a solution to combat fraud on marketplaces, message boards and online stores. Service algorithms based on machine learning can improve the security of online platforms for users and avoid unnecessary costs and expenses due to the actions of fraudsters.

The service developed by the Big Data MTS Scoring team helps to determine fraud at the user authorization stage. Service algorithms analyze impersonal information about patterns of user behavior using data from the MTS ecosystem and other operators, and determine which of them is likely to be unreliable. For example, such features may be data on the duration of the SIM card, the presence of a virtual SIM card, statistics on the number and duration of calls, and another 60 different features.

The anti-fraud service works in conjunction with the online platform scoring system. As a result, this makes it possible to significantly improve the performance and increase the efficiency of searching for fraudsters. The synergy of MTS own analytics and big data allows sites to increase the accuracy of identifying fraudsters by 20%.

File:Aquote1.png
We have already conducted several long successful tests with large customers, in which our system has proven itself well. As of November 2023, we have already fully implemented our anti-fraud service in partner identification systems. Big Data MTS specialists continue to improve algorithms in order to help companies from different industries. For example, as of November 2023, about 50% of all cases of fraud in banks are related to social engineering, and we are working to launch an updated version of the model for more effective protection against fraudsters, "said Victor Kantor, director of the MTS Big Data Center.
File:Aquote2.png