US20250097272
2025-03-20
Electricity
H04L65/1059
The Meeting Visualizer patent application describes a system for creating interactive visualizations during online meetings. These visualizations are generated automatically based on the meeting transcript and other related data. The system determines the meeting's category and suggests a suitable visualization layout. Using a generative language model, it creates interactive visualizations that are then shared with participants to enhance engagement and collaboration.
Online communication platforms are essential for facilitating collaboration. While these platforms support sharing static or interactive content, many participants find it challenging to incorporate interactive elements into meetings. This limitation can reduce engagement and information retention. The proposed system addresses these challenges by simplifying the creation of interactive content, thus encouraging its use in meetings.
The system includes a processor and memory with executable instructions that detect trigger conditions during meetings. Upon detection, it selects a visualization layout based on the meeting category and constructs a prompt for a language model using the meeting transcript. The language model generates visualization information, which is formatted according to the selected layout and presented to participants in an interactive workspace. Participants can interact with and modify this content collaboratively.
The method involves detecting trigger conditions during meetings, analyzing session data to categorize the meeting, and selecting visualization layouts accordingly. Participants receive layout options and choose one for the session. The selected layout, along with session data, is processed by a language model to generate visualization content. This content is then synchronized across participants' devices, allowing real-time collaboration.
This approach enhances participant engagement and information retention by providing an interactive virtual whiteboard where participants can interact using various input methods. The automatic generation and updating of visualizations reduce computing resources by predicting relevant content based on session data. This system offers a starting point for collaboration, avoiding the challenges of creating visualizations from scratch.