Multiuser Session Summarization via Feature Extraction and Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users often miss multiuser communication sessions and lack time to view the entire recorded video, as existing solutions do not provide a concise summary of the most interesting segments.
Innovation Solution
A computer-implemented method that extracts features from multiuser communication session data, scores segments based on user interactions and relationships, and generates a summary including segments of interest, personalized for users, using a social network application, feature extraction module, analyzing module, summarizing engine, and user interface engine.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the entire multiuser communication session is recorded and made available, then users can view the complete content, but users do not have the time to view the entire video
Solution Approach 1:
The patent segments the multiuser communication session into multiple time segments, analyzing each segment's features (audio, video, text) to identify interesting portions. This segmentation allows the system to distinguish between meaningful content and redundant information, enabling users to access only the valuable segments rather than the entire recording.
Solution Approach 2:
The patent extracts specific features from the communication session data (audio features, video features, text features) and uses these extracted features to score and rank segments. This extraction process identifies and isolates the most interesting segments, allowing users to access condensed information without viewing the complete original recording.
2Loss of information
If the summary includes all segments of interest, then the summary is comprehensive, but the summary becomes too long for users to quickly review
Solution Approach 1:
The patent applies local quality by customizing the summary for each user based on their social affinity scores with participants. Different users receive different summaries focused on segments most relevant to their personal interests and relationships, ensuring high-quality, personalized information without unnecessary content.
Solution Approach 2:
The patent changes the parameter of summary length by dynamically adjusting the number and duration of segments included in the summary based on user preferences, social affinity scores, and session features. This allows the summary to be condensed to an optimal length that maintains completeness while fitting user viewing preferences.
3Measurement precision
If the system analyzes all features from all segments, then the segmentation accuracy is high, but the processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary action by pre-defining multiple features to extract (audio, video, text) and pre-establishing scoring criteria before analyzing the communication session. This preparation allows the system to efficiently process and score segments without performing complex analysis from scratch, maintaining accuracy while improving processing speed.
Data Source
AI summary
A system and method for summarizing a multiuser communication session is disclosed. A processing unit receives data about the multiuser communication session. A session management module manages the multiuser communication session. A feature extraction module extracts features from the data. An analyzing module analyzes the features to identify segments of interest for the multiuser communication session including a beginning and an end of each segment of interest and a list of participants in each segment of interest. A summarizing engine generates a summary for the multiuser communication session including at least one segment of interest.