Real-Time Natural Language Summaries for Linked Meeting Actions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing meeting technologies fail to efficiently detect and summarize action items, their relations, and solutions in real-time, making it difficult for late attendees to understand meeting progress and their responsibilities.
Innovation Solution
A system utilizing Natural Language Processing (NLP) and Machine Learning (ML) models to analyze meeting conversations, identify action items, their relations, and solutions, and generate real-time summaries for attendees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual tracking of action items is used, then completeness of task tracking is improved, but time consumption and efficiency deteriorate
Solution Approach 1:
The system automatically extracts and tracks action items from meeting transcripts without requiring manual intervention. The NLP model autonomously identifies tasks, assigns them to participants, and updates their status, enabling the system to serve itself in tracking tasks while eliminating manual time consumption.
Solution Approach 2:
The patent replaces manual mechanical tracking methods with an automated NLP-based system. The machine learning model processes meeting transcripts and automatically extracts action items, substituting human manual effort with computational processing to achieve both completeness and efficiency.
2Loss of information
If detailed meeting transcripts are provided to late attendees, then information completeness is improved, but ease of understanding deteriorates
Solution Approach 1:
The system extracts only the essential action items from complete meeting transcripts and presents them to late attendees. By separating and extracting the most critical information (tasks, assignments, and status) from the full transcript, the system maintains information completeness while dramatically improving ease of understanding.
Solution Approach 2:
The patent segments the complete meeting transcript into distinct action items with specific attributes (task description, assignee, status). This segmentation transforms a large block of text into structured, digestible units that are easier for late attendees to understand while preserving all essential information.
3Productivity
If real-time action item detection is implemented, then productivity is improved, but system complexity increases
Solution Approach 1:
The patent introduces an NLP processing layer as an intermediary between meeting transcripts and action item extraction. This intermediary layer handles the complexity of real-time text analysis, enabling productive real-time processing while shielding the rest of the system from computational complexity through modular architecture.
Data Source
AI summary
A system receives messaging, video and/or audio input streams including dialogue spoken by users at a group meeting. From these inputs, the system obtains single or multiple interaction records including natural language text memorializing content spoken by each speaker at a meeting, analyzes the content, and identifies single or multiple action item tasks in the interaction records. The system then generates summaries indicating the action item tasks for the users. From the dialogue content, the system further detects whether each action item is addressed, and whether the action item for a user has a solution, or not. The system further detects whether one action item is a precondition for resolving another action item by the user or in conjunction with another user. Using a pre-configured template, the system generates action item summaries, any associated solution, and any relationship or precondition between action items and presents the summary to a user.