RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

WaveAccess Antifraud system

Product
The name of the base system (platform): Artificial intelligence (AI, Artificial intelligence, AI)
Developers: WaveAccess (Veyvaksess)
Date of the premiere of the system: 2017/10/25
Branches: Telecommunication and communication
Technology: Cybersecurity - Fraud detection system (fraud)

Antifraud system — the solution created based on algorithms of machine learning and a fuzzy logic for protection of telecommunication companies against the attacks of swindlers.

2017: Release

The WaveAccess company provided on October 25, 2017 an antifraud system, the protection of the telecom companies against swindlers designed to provide. At the heart of the solution — algorithms of machine learning and a fuzzy logic.


Prerequisites

It is easy to confuse the fraud facts on networks of the operator with actions of the ordinary user, for example, with an error of input or access from the new device. At the same time swindlers constantly invent cracking methods, and standard software tools of protection cease to work. But a method to prevent the attack is — a system should "understand" as far as the event is similar to threat, noted in WaveAccess.

How it works

For this purpose specialists of WaveAccess made classification of the most widespread threats — only 16 types — and trained the module at a set of these data, having provided reactions to each threat. For example, bots sorters of passwords a system automatically asks to pass a test for character recognition ("kapch"), the same occurs also at access attempt in one account from several IP. The algorithm of determination is provided for each threat, and it is not always a feedback request from the user. Sometimes a system just blocks suspicious IP.

Nagruzoustoychivost

The module of protection was "trained" to distinguish the attacks of hackers from normal behavior of users even if in a system hundreds of thousands of users at the same time will become authorized. The Nagruzoustoychivost was tested by means of the tool which determined correctness of work of an algorithm by the set parameters and imitated loading — created a necessary data stream. Attention was paid also to the interface which was tested manually on the basis of the provided scenarios.

Features

To trace the events, the module with the user interface is developed for system administrators. It opens access to security logs (magazines in which events are fixed), allows to configure system settings.

Scopes of application

File:Aquote1.png
The problem of a fraud is relevant for banks, payment systems, a telecom. Mobile operators annually lose billions because of fraud. Access attempt on counterfeit IP, plunder of the credit card number, password guessing and other attacks — cannot be guessed what will be thought up by swindlers. Therefore it is impossible simply "to deliver an antivirus", special tools which study behavior of users are necessary and compare it to examples of suspicious actions. Our command developed and trained the module which is designed to save data of the telecom companies, to increase safety of transactions in a billing system. We are ready to adapt the solution and for other industries — Alexander Azarov, Senior Vice President of Engineering in WaveAccess emphasized.
File:Aquote2.png