Fraud Detection via Call Characteristic Distribution Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fraud detection methods in call-based advertising rely heavily on caller ID, which can be easily circumvented by fraudsters using temporary numbers or fake IDs, necessitating a more robust and ID-independent solution.
Innovation Solution
A system and method that analyzes call characteristics such as call length, interaction patterns, and time to identify distributions indicative of fraudulent activity, using statistical processing and machine learning algorithms, without relying on caller ID or call recording/transcription, to detect and mitigate fraudulent calls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If caller ID based fraud detection is used, then fraud detection capability is improved, but reliability deteriorates because fraudsters can easily circumvent by using temporary numbers or fake IDs
Solution Approach 1:
The patent transitions from using caller ID parameters to analyzing call characteristics parameters such as call duration, time of day, day of week, and interaction patterns. This parameter change makes fraud detection more reliable because these behavioral parameters are harder for fraudsters to manipulate compared to caller ID information.
Solution Approach 2:
The patent replaces the mechanical caller ID verification system with a statistical analysis system that uses machine learning algorithms. Instead of relying on static ID matching, the system uses probabilistic models to analyze call patterns and predict fraud likelihood, providing more robust detection capability.
2Measurement precision
If call recording or transcription is used for analysis, then measurement precision is improved, but loss of information increases due to privacy concerns about personal data
Solution Approach 1:
The patent extracts only the necessary call characteristics (duration, timing, interaction patterns) from the call data without retaining the full call content. By taking out only the essential metadata needed for fraud detection and discarding the rest, the system achieves accurate fraud detection while minimizing privacy information loss.
Solution Approach 2:
The system uses temporary, disposable call characteristic data that is analyzed immediately and then discarded. Instead of storing permanent call recordings or transcriptions, the system processes ephemeral metadata and deletes it after analysis, reducing privacy risks while maintaining detection effectiveness.
Data Source
AI summary
A system and method for monitoring telephone calls to detect fraudulent activity and take corrective action is described. The system receives a group of telephone calls having associated call characteristics and analyzes the group of telephone calls to identify and store a first set of distributions of call characteristics that are indicative of normal activity, fraudulent activity, or indeterminate activity. The system receives one or more subsequent telephone calls to be analyzed. The system analyzes the received one or more telephone calls to identify a second set of distributions of call characteristics associated with the received telephone call. The system then compares the second set of distributions of call characteristics to the stored first set of distributions of call characteristics to assess a probability that the one or more received telephone calls represents normal, fraudulent, or indeterminate activity. If the assessed probability of fraudulent activity exceeds a threshold, the system takes appropriate corrective action, such a flagging the fraudulent call or withholding a financial incentive associated with the fraudulent call.