User Reputation Graph for Content Quality Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulties in finding desirable content among vast amounts of user-submitted content on social collaborative networks, as unscrupulous individuals may post spam or duplicate content, making it hard to locate legitimate content.
Innovation Solution
A system that evaluates items based on user reputation information by using a preference system where users can submit, digg, bury, and comment on content, with a user reputation engine assigning scores to users and an item scoring engine determining content promotion based on these interactions, including the construction of graphs from sharing event logs to compute user reputation scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If user-submitted content is accepted to increase platform content volume, then content variety and user engagement are improved, but spam and duplicate content increase making legitimate content harder to locate
Solution Approach 1:
The system implements a feedback mechanism where users can digg (upvote) or bury (downvote) content, and this feedback is used to calculate user reputation scores. Users with higher reputation scores are trusted more, and their content is given higher visibility. This feedback loop enables the system to automatically filter spam and duplicate content while promoting high-quality user-submitted content, resolving the contradiction between content volume and quality.
Solution Approach 2:
The patent replaces manual content moderation with an automated reputation-based filtering system. Instead of requiring human reviewers to manually assess each piece of content, the system uses computational algorithms that calculate user reputation scores based on voting patterns and content performance data. This automated mechanism efficiently handles large volumes of user-submitted content while maintaining quality standards.
2Reliability
If manual content moderation is implemented to filter spam and duplicate content, then content quality is improved, but system complexity and operational burden increase
Solution Approach 1:
The system enables users to self-regulate content quality through the digg/bury mechanism. Users actively participate in quality control by voting on content, and the system automatically aggregates these votes to generate reputation scores. This self-service approach distributes the moderation function across the user base rather than requiring a centralized complex moderation system, reducing operational burden while maintaining content quality.
3Measurement precision
If user reputation scores are calculated using complex graph algorithms, then content evaluation accuracy is improved, but computational time and processing resources increase
Solution Approach 1:
The system performs preliminary calculations by pre-computing and storing user reputation scores and content metrics. Rather than calculating reputations in real-time during content evaluation, the system prepares and stores these values beforehand, allowing for faster subsequent content ranking and filtering operations. This preliminary action reduces computational time during critical content evaluation phases while maintaining accurate reputation-based filtering.
Data Source
AI summary
Evaluating an item based on user reputation information is disclosed. Sharing event data is received. A graph is materialized from the received sharing event data. User reputation scores are determined based at least in part on the graph. A score is assigned to the item based at least in part on the user reputation scores of users associated with the item.