Online Meeting Diarization for Group Dynamics Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing online meetings fail to provide insights into group dynamics, group behavior, and meeting efficiency, particularly in substituting for the deficiency of non-verbal communication.
Innovation Solution
A system and method for dynamically generating and analyzing metadata from online meetings using machine learning algorithms to extract insights on participant behavior, group dynamics, and meeting effectiveness, including diarization to segment audio data and identify individual speakers, with real-time visualization and recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing methods for analyzing online meetings are used, then basic meeting data can be captured, but insights into group dynamics, group behavior, and meeting efficiency cannot be provided
Solution Approach 1:
The audio stream is segmented into individual speaker segments through diarization, separating the mixed audio signal into distinct participant contributions. This enables granular analysis of individual speaking patterns while maintaining the ability to analyze group dynamics through aggregation of segment-level features.
Solution Approach 2:
Metadata acts as an intermediary layer between raw audio streams and high-level insights. The system extracts metadata features (speaking time, pause duration, interruption patterns) that bridge the gap between basic audio capture and complex group behavior analysis, enabling efficient computation of group dynamics metrics.
2Adaptability or versatility
If non-verbal communication is removed in online meetings, then geographical dispersion is enabled, but ability to experience and participate in non-verbal communication is reduced
Solution Approach 1:
The system substitutes acoustic signals for non-verbal communication cues. By analyzing audio characteristics (pause durations, interruption patterns, speaking tempo, pitch variations), the system recovers proxies for non-verbal behaviors such as engagement level, agreement, disagreement, and turn-taking dynamics that are typically conveyed through body language and facial expressions.
3Productivity
If automated analysis is implemented, then meeting efficiency insights can be provided, but processing complexity increases
Solution Approach 1:
The system performs preliminary extraction of metadata features during the meeting, calculating speaking time, pause durations, and turn-taking patterns in real-time. This preliminary processing organizes raw audio data into structured metrics before final group dynamics analysis, reducing computational complexity of the overall system by pre-computing intermediate representations.
Solution Approach 2:
The diarization system automatically identifies speakers and segments audio without manual intervention, and the analysis system automatically computes group dynamics metrics from the segmented data. This self-service automation eliminates the need for manual transcription and analysis, providing meeting efficiency insights while managing processing complexity through algorithmic automation.
Data Source
AI summary
A system for dynamically generating and analyzing metadata for online meetings is provided. The system is programmed to: a) receive at least one stream of at least one of audio and video of an online meeting, wherein the at least one stream includes one or more participants participating in the online meeting; b) extract a plurality of metadata from the at least one stream; c) perform diarization on the at least one stream and the plurality of metadata the at least one stream to generate diarization information, wherein the diarization information includes information about participation for the one or more participants in the online meeting; d) analyze the diarization information to calculate one or more key performance indicators; and e) generate visualization of the key performance indicators to be displayed to one or more participants in the online meeting.