Truth Discovery via Source Dependency Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing truth discovery methods fail to accurately distinguish true facts from conflicting data sources due to the complexity introduced by source dependency, where data is copied between sources, leading to biased voting algorithms and incorrect accuracy assessments.
Innovation Solution
A methodology that estimates dependency between sources by calculating shared and false shared data object values, determining dependency probabilities, and generating dependency graphs to assess truth counts, while considering source accuracy and dependency polarity to improve the reliability of truth discovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If voting algorithms are used to determine truth from multiple sources, then truth discovery can be performed, but source dependency causes biased results and incorrect accuracy assessments
Solution Approach 1:
The patent introduces an intermediary dependency detection mechanism that mediates between multiple data sources before applying voting algorithms. This intermediary layer identifies and quantifies source dependencies, allowing the system to adjust or weight votes accordingly, thereby eliminating the bias that would otherwise corrupt the truth discovery process
Solution Approach 2:
The system implements feedback by continuously monitoring source relationships and using detected dependency information to refine the truth discovery process. The dependency detection results feed back into the voting mechanism to adjust source credibility scores, creating a self-correcting system that accounts for source interdependencies
2Quantity of substance
If multiple data sources are integrated, then data completeness improves, but determining actual truth becomes more complex due to source dependency
Solution Approach 1:
The patent segments the truth determination process into distinct modules: a dependency detection module that analyzes source relationships, and a truth discovery module that applies voting algorithms. This segmentation allows the system to handle multiple data sources systematically by first identifying dependencies and then applying appropriate weighting, reducing the overall complexity
Solution Approach 2:
The system changes parameters by introducing dependency scores and credibility weights as new variables. These parameters transform the raw voting process into a weighted evaluation system, where the complexity is managed through mathematical modeling rather than ad-hoc analysis, making the process more tractable
3Ease of manufacture
If source accuracy is assessed using traditional methods, then simplicity is maintained, but accuracy assessments are incorrect due to unaccounted source dependency
Solution Approach 1:
The patent applies preliminary action by detecting and quantifying source dependencies before the accuracy assessment is performed. This preliminary dependency analysis prepares the system by establishing source relationship metrics in advance, allowing subsequent accuracy assessments to incorporate this information without adding significant complexity to the overall process
Data Source
AI summary
A method and system for truth discovery may implement a methodology that accounts for accuracy of sources and dependency between sources. The methodology may be based on Bayesian probability calculus for determining which data object values published by sources are likely to be true. The method may be recursive with respect to dependency, accuracy, and actual truth discovery for a plurality of sources.