Data Accuracy System Using Trustworthiness Weights
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Financial institutions face challenges in accurately verifying data due to varying degrees of trustworthiness across different data sources, leading to unreliable risk determinations and potential exclusion of useful information from untrusted sources.
Innovation Solution
A system and method for assessing data record accuracy by calculating an accuracy score that combines trustworthiness weights and matching scores from multiple data sources, including both trusted and untrusted sources, to evaluate the reliability of data fields and records.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple data sources with varying trustworthiness are used to verify data, then the quantity of data sources increases and more information becomes available, but the reliability of risk determination deteriorates due to inconsistent trustworthiness levels
Solution Approach 1:
The patent applies local quality by assigning different trustworthiness weights to different data sources based on their individual reliability characteristics. Each data source is evaluated and assigned a specific weight that reflects its local quality or trustworthiness level, allowing the system to differentiate between high-quality and low-quality sources rather than treating all sources uniformly.
Solution Approach 2:
The patent changes the parameter of trustworthiness by introducing a quantitative weight system. The trustworthiness of each data source is transformed into a numerical weight that can be mathematically applied to the data from that source. This parameter change allows for systematic integration of multiple sources with varying reliability by adjusting their influence on the final risk determination through weighted aggregation.
2Reliability
If data sources with lower trustworthiness are excluded from risk determination, then the reliability of risk determination improves, but the quantity of usable information decreases
Solution Approach 1:
The patent converts the potential harm of using untrusted data sources into a benefit by implementing a weighting system. Data sources with lower trustworthiness are not completely excluded but are instead assigned lower weights, allowing their information to contribute to the analysis in proportion to their reliability. This transforms what would be harmful noise into potentially useful information that is appropriately discounted.
Solution Approach 2:
The patent applies partial action by including data from all sources rather than completely excluding low-trustworthiness sources. The inclusion is partial in the sense that the influence of each source is modulated by its trustworthiness weight, allowing the system to benefit from information across the full spectrum of sources while mitigating the risk from less reliable ones through proportional weighting.
3Measurement precision
If only highly trusted data sources are used for data verification, then the trustworthiness of data assessment improves, but the quantity of data sources available for verification decreases
Solution Approach 1:
The patent creates a composite assessment by combining data from multiple sources with different trustworthiness levels, similar to creating composite materials. Each data source contributes a different 'component' to the final assessment, with the trustworthiness weights acting as proportions that determine the contribution of each component. This composite approach allows the system to leverage the strengths of multiple sources while accounting for their individual limitations.
Data Source
AI summary
Data from a plurality of data sources is provided to a multi-source data management system, which stores the data and provides it to a data accuracy system for purposes of assessing the accuracy of data records and the individual fields within data records. Data accuracy scores may be stored at the data management system with the data records to which they pertain. Accuracy scores may be periodically recalculated and monitored, and alerts provided if an accuracy score changes a predetermined amount over a given period of time. Also, data records may be provided by a data user for accuracy assessment, using other data records stored at the multi-sourced data management system.