With the development and popularization of telecom technology, global telecom operators are now facing increasingly serious fraud problems which have already caused huge revenue loss to them and obstructed their normal operation. Therefore, it is significant to effectively detect and prevent fraud activities in telecom industry.
Based on our professional knowledge in telecom industry and up-to-date technology, ZSmart Fraud Management System (FMS) is designed to detect, analyze and handle various fraud activities in telecom industry in order to lower risk, reduce revenue loss for operators, protect telecom customer's benefit and guarantee smooth telecom operation.
Excellent ExpandabilityZSmart FMS System supports the configuration of fraud scenarios to meet new fraud detection needs. As the business grows up, ZSmart FMS can be expanded horizontally to enhance the processing capabilities, including data extraction, data analysis and pre-processing and fraud detection.
- Dynamic Extending of Analysis ObjectZSmart FMS can expand fraud-monitoring range through adding analysis objects.
- Dynamic Extending of Reference Data SourceIn ZSmart FMS, the Reference Objects specifies the data sources used during the process of fraud detection. By defining reference objects uniformly in Reference Object Definition module, new data sources can be easily introduced in fraud detection patterns. For example, the Reference Object ‘holiday’ can be specified in the pattern to diminish the false alarms on the higher usage in a holiday.
- Fraud Detection ModuleProvide abundant fraud scenario library to realize rapid deployment and flexible expansion of fraud detection modules.
Advanced Detection Concept
- Scenario Based Fraud DetectionZSmart FMS supports flexible fraud detection engine and can detect the abnormal behavior and fraud on basis of fraud scenarios.
- Detection based on AI technique ZSmart FMS support detection based on AI technique such as decision tree and neural network algorithms to carry out analysis, learning and training toward fraudsters’ history consumption data, extracting hidden business rules to generate different fraud analysis models (i.e. Fraud Pattern).
- Auto Threshold EngineThe Auto threshold engine calculates the threshold automatically for a subscriber depending on the usage pattern and the call duration/call times.
- Statistical Algorithm ModelsStatistical Algorithm models are supported for generating individual subscriber’s usage pattern.
- Fraud Case Library Provide fraud case library for reference and study.
Workflow-based ProtectionZSmart FMS can help operators to realize fraud protection based on workflow. Different workflows are specified for different fraud scenarios, and therefore, once the alarm is detected as a fraud, the corresponding workflow can be started to coordinate various systems and departments involved to handle the fraud behavior together.
- Protect revenues and minimize losses
- Protects reputation of company
- Protect customers from criminal activities
- Ensures customers confidenceZSmart FMS