Alert Manager Detecting Link Mismatches via Predictive Timing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional systems fail to effectively alert users about potentially fraudulent messages, particularly those containing links from suspicious sources, as they often rely on visual cues that may not be timely or reliable, allowing phishing attempts to deceive users.

Innovation Solution

The Alert Manager system detects message receipt, compares the message's source and link destination addresses for discrepancies, and uses historical user browsing behavior to predict when the user is likely to interact with the link, generating a visual alert feature at the moment of predicted interaction to warn the user about potential fraud.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If visual cues are used to alert users about potentially fraudulent messages, then users can be warned about phishing attempts, but the alerts may not be timely or reliable enough to prevent user deception

Engineering Contradiction:
Improvealert reliabilityVSAvoidresponse time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary analysis by comparing the message source address with the link destination address before the user interacts with the link. This advance detection allows the system to prepare and deliver alerts at the precise moment of predicted user interaction, ensuring both timeliness and reliability of the warning.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses historical user browsing behavior to generate predictions about future user actions. This feedback loop continuously learns from user patterns to improve the timing and accuracy of alerts, making the warning system progressively more reliable and timely based on actual user interaction data.

Inventive Principle:
Principle #23Feedback

2Reliability

If the system compares message source addresses with link destination addresses to detect fraud, then phishing attempts can be identified, but the system complexity increases

Engineering Contradiction:
Improvefraud detection accuracyVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system extracts only the critical elements needed for fraud detection: the message source address and the link destination address. By focusing on these specific extracted elements rather than analyzing the entire message content, the system achieves reliable fraud detection while maintaining relatively simple processing logic.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system creates a simplified representation by comparing two key address strings (source and destination). This copying approach uses basic string comparison operations rather than complex analysis, achieving effective fraud detection through a straightforward methodology that minimizes system complexity.

Inventive Principle:
Principle #26Copying

3Loss of time

If historical user browsing behavior is analyzed to predict user interaction timing, then alerts can be delivered at the optimal moment, but the processing requirements and system complexity increase

Engineering Contradiction:
Improvealert timing precisionVSAvoidprocessing complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The system performs partial analysis by focusing only on the timing aspect of user behavior rather than comprehensively analyzing all user interaction patterns. By extracting and analyzing only the temporal elements of browsing behavior, the system achieves precise alert timing while avoiding the complexity of full behavioral analysis.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3394765B1Detecting hidden link mismatches with spoofed metadata
Publication Date: 2021.06.16 EBAY INC
  • EP3394765B1 patent drawingFigure 1
  • EP3394765B1 patent drawingFigure 2
  • EP3394765B1 patent drawingFigure 3

AI summary

Example methods and systems directed to an Alert Manager are described. According to various embodiments, the Alert Manager detects receipt of a message. The message includes a selectable functionality for accessing an external resource and message data indicative of a source of the message. The Alert Manager predicts when a recipient of the message will interact with the selectable functionality. Prior to the recipients predicted interaction with the selectable functionality, the Alert Manager generates a message alert feature based on a degree of a difference between the external resource and the source of the message.