API Proposal_Fraud Hotzone Report.md

July 30, 2025 · View on GitHub

FieldDescription
API family nameFraud Hotzone Report
API family ownerChunghwa Telecom
API summaryThis API analyzes recent SMS activity patterns for a suspected fraud victim's phone number. It detects anomalies based on adaptive AI models, taking into account fraud-prone countries or regions (fraud hotzones) specified by the user. The result helps determine whether the observed behavior indicates potential high-risk fraudulent targeting within the given lookback period.
Technical viabilityThis API utilizes call and SMS activity data retrieved from mobile core network sources (e.g., IMS information) and applies our custom-designed adaptive AI algorithms, along with other detection methods, to identify abnormal communication behaviors associated with fraud hotspots and determine the dates on which such anomalies occurred.
Commercial viabilityIn current situation many scam organizations use phone call and SMS interactions to carry out fraudulent activities.

Challenges in Anti-Fraud Application Design:
•Accurately identifying whether a target number is high risk is difficult.
•Anti-fraud applications can only make educated guesses based on limited scenarios.

We propose the Communication Risk Check API, which provides anti-fraud indicators of target phone numbers. This enables anti-fraud applications to assess the risk from two perspectives: detecting potential scammers and protecting potential victims.
YAML code available?NO.
Validated in lab/productive environments?NO.
Validated with real customers?NO.
Validated with operators?NO.
Supporters in API Backlog Working GroupChunghwa Telecom