Meeting Security Labeling via ML Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Meetings and recordings are prone to data leaks due to eavesdropping, visual spying, and accidental sharing of sensitive information, especially in hybrid work modes where remote conferencing tools are used.
Innovation Solution
The system predicts meeting security levels, assesses security in recorded meetings, and provides recommendations to modify meeting access both before and after the meeting, using machine learning algorithms to apply security labels and ensure appropriate access controls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If meeting recordings are shared with attendees, then information accessibility is improved, but data security deteriorates due to potential leaks of sensitive information
Solution Approach 1:
The patent applies local quality by assigning different security labels (public, internal, confidential, restricted) to different portions of meeting content based on its sensitivity. This allows selective access control where different users can access different segments of the recording according to their authorization level, thus maintaining security while preserving accessibility for authorized users.
Solution Approach 2:
The meeting recording is segmented into multiple portions with different security classifications. The system divides the continuous recording stream into discrete segments that can be independently controlled and distributed, allowing the organization to share accessible portions while protecting sensitive segments from unauthorized access.
2Object-affected harmful factors
If security labels are applied to meeting content, then data security is improved, but device complexity increases due to additional processing requirements
Solution Approach 1:
The system performs preliminary action by automatically analyzing meeting content during or immediately after the meeting to identify sensitive information and apply appropriate security labels. This pre-processing approach eliminates the need for manual review and classification, reducing operational complexity while maintaining high security standards.
Solution Approach 2:
The meeting content analysis system operates autonomously to detect sensitive information and apply security labels without requiring manual intervention. The system self-services by using machine learning models to automatically classify content, thereby reducing the complexity burden on users while ensuring consistent security application.
3Measurement precision
If machine learning algorithms are used to analyze meeting content, then security assessment accuracy is improved, but productivity decreases due to increased processing time
Solution Approach 1:
The system applies partial action by focusing machine learning analysis only on potentially sensitive portions of meeting content rather than processing the entire recording uniformly. This selective approach maintains high security assessment accuracy for critical segments while reducing overall processing time and computational resources required.
Data Source
AI summary
Disclosed are various embodiments for preventing unauthorized access to materials and topics for meeting invitations and meeting recording. Meeting data is identified for a meeting invitation or a meeting recording. The meeting data includes a set of participants, a time slot, and meeting location. A meeting security recommendation is generated to include a security label based on output data from a meeting security service machine learning engine that takes the meeting data as input. The meeting security recommendation is transmitted to a client device. A meeting security action is implemented using security recommendation feedback received from the client device.