Scam Call Detection System Using Dynamic Weight Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current call validation methods are burdensome and not fast enough to maintain real-time scam call management, as they fail to efficiently identify and filter out undesired calls from various sources, including spam, scam, and robocalls.
Innovation Solution
A method that involves identifying call parameters, assigning scores to calls, categorizing them as scam or legitimate, and adjusting weights for call parameters based on deviation from a target percentage to determine whether to permit or block calls, using a processor and non-transitory computer-readable storage medium to execute these processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current call validation methods are used to identify legitimate calls, then call authenticity can be verified, but the process becomes burdensome and too slow for real-time scam call management
Solution Approach 1:
The system performs preliminary scoring of calls based on multiple parameters (caller ID reputation, call frequency, time of day, day of week, call duration) before the actual validation decision is made. This preliminary assessment filters out obvious scam calls quickly, allowing only ambiguous cases to undergo more thorough validation, thus maintaining reliability while improving processing speed.
Solution Approach 2:
The call validation process is segmented into multiple independent scoring components, each evaluating different aspects of the call (caller reputation, frequency patterns, temporal characteristics). These segmented scores are then aggregated to form an overall scam likelihood assessment, enabling parallel processing of multiple validation criteria simultaneously.
2Measurement precision
If multiple call parameters are analyzed to accurately categorize calls as scam or legitimate, then scam call identification accuracy improves, but system complexity increases
Solution Approach 1:
The system dynamically adjusts the weights assigned to different call parameters based on observed patterns and performance metrics. For example, if caller ID reputation proves to be a strong indicator of scam calls, its weight in the overall scoring algorithm is increased. This allows the system to maintain high detection accuracy while adapting to changing scam patterns without requiring complete system redesign.
Solution Approach 2:
The system incorporates feedback loops where the outcomes of call categorizations are fed back into the scoring algorithm to refine parameter weights and thresholds. Successful scam call detections reinforce the effectiveness of certain parameter combinations, while false positives trigger adjustments to reduce their impact, continuously improving accuracy without linearly increasing complexity.
3Adaptability or versatility
If dynamic weight adjustment for call parameters is implemented to adapt to changing scam patterns, then scam call detection effectiveness improves, but computational overhead increases
Solution Approach 1:
The system performs dynamic weight adjustment periodically rather than continuously, updating parameter weights at scheduled intervals or after processing a certain number of calls. This periodic recalibration maintains adaptability to changing scam patterns while avoiding the continuous computational overhead of real-time weight adjustment for every single call.
Solution Approach 2:
The system implements dynamic weighting where parameter importance can change based on current scam trends, but this dynamics is constrained to predefined adjustment ranges and algorithms. The adaptability mechanism focuses computational resources on adjusting only the most critical parameters that show significant changes in scam patterns, rather than recalculating all weights equally.
Data Source
AI summary
One example method of operation may include identifying one or more call parameters associated with each of a number of calls received over a fixed period of time, assigning scores to each of the calls based on the one or more identified call parameters for each of the plurality of calls, assigning one or more of the calls to a scam call category based on the assigned scores, and responsive to the assigning of the one or more of the calls to a scam call category, determining whether a number of remaining calls of the calls, which are not assigned to the scam call category, have increased or decreased beyond a deviation margin of a target percentage of calls.