Video Newsworthiness Detection via Knowledge Base Entity Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Social networks face challenges in determining which videos are newsworthy among the vast amount of content uploaded every minute, leading to difficulties in sending relevant notifications to users without overwhelming them with intrusive messages.
Innovation Solution
A method that uses a knowledge base to attribute newsworthiness to videos by identifying embedded videos in online content, recording their IDs, and determining videos with the highest occurrences within a knowledge base entity, while considering language, geographic categorization, and user affinity scores to tailor notifications to users likely to find them important.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If notifications are sent for all uploaded videos, then users receive comprehensive information about new content, but users are overwhelmed with intrusive messages
Solution Approach 1:
The patent applies local quality by differentiating notification treatment based on video characteristics. Instead of uniform notification for all videos, the system identifies newsworthy videos through knowledge base entity matching and sends targeted notifications only for those, while excluding routine or non-news content from notifications.
Solution Approach 2:
The system changes the parameter of notification selection from binary (all or none) to selective based on newsworthiness criteria. By introducing parameters such as knowledge base entity occurrence frequency, language matching, and geographic categorization, the system transforms notification delivery from comprehensive to precision-targeted.
2Object-affected harmful factors
If notifications are reduced to avoid intrusiveness, then user experience improves, but relevant newsworthy videos may be missed
Solution Approach 1:
The patent introduces knowledge base entities as intermediaries between videos and users. Instead of direct notification based on simple upload detection, the system uses knowledge base entity matching as a mediator to filter and identify which videos represent genuine newsworthy events, ensuring neither over-notification nor missed important content.
Solution Approach 2:
The system replaces mechanical video upload detection with semantic analysis through knowledge base entity recognition. This substitution enables intelligent differentiation between routine content and newsworthy content, allowing accurate identification of videos that truly merit user notification.
3Loss of information
If all videos are analyzed for notification eligibility, then comprehensive coverage is achieved, but computing resources are consumed
Solution Approach 1:
The patent extracts only the essential特征 for newsworthiness determination from videos by matching against knowledge base entities. Instead of analyzing all video content comprehensively, the system extracts key identifying information (titles, descriptions, metadata) and compares these against the knowledge base, significantly reducing computational requirements while maintaining coverage.
Solution Approach 2:
The system applies partial action by performing knowledge base entity matching only on videos that meet initial criteria rather than analyzing every uploaded video in depth. This selective approach processes only the necessary subset of videos for notification eligibility, conserving computing resources while achieving adequate coverage.
4Loss of information
If notifications are sent to all users for newsworthy videos, then maximum awareness is achieved, but resource efficiency decreases
Solution Approach 1:
The patent applies local quality by differentiating user notification eligibility based on individual user characteristics such as language preferences, geographic location, and interaction history. Instead of uniform notification to all users, the system tailors notification delivery to specific user segments, reducing overall notification volume while maintaining awareness for relevant users.
Solution Approach 2:
The system uses feedback from user interaction history with knowledge base entities to refine notification targeting. By analyzing past user behavior patterns, the system identifies users most likely to engage with newsworthy content in their domain, optimizing notification delivery to high-probability users and reducing waste on low-probability recipients.
Data Source
AI summary
Implementations disclose determining which videos are newsworthy events. A method includes identifying, based on data extracted from online content, a knowledge base (KB) entity corresponding to the online content, identifying an embedded video in the online content, recording an online content identifier (ID) of the online content and a video ID of the embedded video as an entry for the identified KB entity in a KB reference data store, determining at least one video ID that occurs more than a threshold number of times in the entry for the identified KB entity in the KB reference data store, and identifying a resulting video corresponding to one of the at least one video IDs as a candidate for a notification to one or more users.