Real-Time Compliance Analysis Using Dependency Parse Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing compliance mechanisms for digital messaging are ineffective as they rely on employees to self-police and typically analyze communications after the fact, leading to false positives and inability to prevent non-compliant messages from being sent.
Innovation Solution
A system that captures keystrokes and analyzes digital messages in real-time to identify non-compliant expressions by using concept flags, relation patterns, and dependency parse graphs, and prevents the transmission of such messages by blocking them at the endpoint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If real-time message analysis is implemented, then compliance detection accuracy is improved, but system complexity increases
Solution Approach 1:
The system segments message analysis into distinct components: keystroke capture module, real-time analysis module, and blocking module. Each module handles a specific aspect of compliance detection, allowing the complex analysis task to be divided into manageable parts that can be processed independently and efficiently.
Solution Approach 2:
The patent introduces an intermediary analysis layer between message composition and transmission. This intermediary module intercepts messages during composition, analyzes them against compliance policies, and determines whether to allow transmission. This mediator approach enables sophisticated analysis without requiring changes to the underlying messaging infrastructure.
2Reliability
If real-time blocking of non-compliant messages is implemented, then compliance violation prevention is improved, but false positives increase
Solution Approach 1:
The system performs preliminary analysis of messages during composition before transmission. By analyzing messages in advance and providing feedback to users during drafting, the system allows users to correct potential compliance issues before sending. This preliminary intervention reduces false positives by giving users context and opportunity to adjust their messages.
Solution Approach 2:
The patent implements a feedback mechanism that provides real-time notifications to users when their message composition may violate compliance policies. This feedback includes specific guidance about what makes the message problematic and suggestions for modification. Users can review feedback and adjust their messages before sending, reducing false positives while maintaining high compliance detection accuracy.
3Adaptability or versatility
If comprehensive message monitoring is implemented, then compliance detection capability is improved, but user privacy concerns increase
Solution Approach 1:
The system applies compliance monitoring with different levels of intensity based on local context. It monitors messages differently depending on the communication channel, recipient, and content sensitivity. This localized approach allows comprehensive compliance detection where needed while minimizing monitoring in contexts where it is less critical, thereby addressing privacy concerns.
Solution Approach 2:
The patent implements a system where users have control over their own message monitoring through configurable settings. Users can adjust monitoring sensitivity, review detected issues before blocking occurs, and manage their own compliance preferences. This self-service approach empowers users to balance compliance detection with their own privacy preferences.
Data Source
AI summary
A method for analyzing a digital message for presence of a non-compliant expression within the digital message is disclosed. The digital message may be created using a computing device including a text input hardware device and may contain at least one segment. The method includes capturing keystrokes at the text input hardware device, analyzing the keystrokes to capture the digital message including an end of the segment, and assessing the segment for presence of a concept flag in at the segment, wherein the concept flag is associated with the non-compliant expression. Assessing may include obtaining a relation pattern for the concept flag, generating a dependency parse graph for the segment, and comparing the dependency parse graph to the relation pattern. If at least a portion of the dependency parse graph matches the relation pattern, then concluding that the segment contains the concept flag associated with the non-compliant expression.


