Record Linkage Using Temporal Pattern Weights for Accurate Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing record linkage methods, such as the Fellegi-Sunter approach, fail to accurately match patient records due to issues like false-negatives and false-positives, especially with spelling errors, missing data, and inconsistent identifiers, leading to incomplete data sharing and potential errors in healthcare.
Innovation Solution
Incorporating power-spectrum-based temporal pattern-specific weights into record linkage methods, using Bayesian power spectra analysis to calculate value-specific weights and timeseries-derived likelihood ratios, enhancing matching accuracy by leveraging time-oriented information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Fellegi-Sunter record linkage method is used, then matching process is simple and fast, but accuracy is poor due to false-negatives and false-positives
Solution Approach 1:
The patent changes the parameters of the record linkage method by introducing temporal pattern-specific weights and power spectrum analysis. Instead of using uniform field-specific weights as in Fellegi-Sunter, the system calculates weights based on the power spectrum of temporal patterns, transforming the weighting mechanism from static to dynamic and from uniform to adaptive, thereby improving accuracy while maintaining processing efficiency
Solution Approach 2:
The patent introduces dynamic weighting where the weights are not fixed but are calculated based on the power spectrum characteristics of temporal patterns. The system dynamically adjusts the importance of different temporal patterns based on their spectral properties, allowing the linkage algorithm to adapt to different data characteristics and improve accuracy without sacrificing speed
2Device complexity
If Fellegi-Sunter method is used, then computational complexity is low, but false-negatives increase due to spelling errors and missing data
Solution Approach 1:
The patent incorporates feedback mechanisms through power spectrum analysis that continuously evaluates temporal patterns and adjusts weights accordingly. The system analyzes the spectral characteristics of temporal data and uses this feedback to refine the matching process, thereby reducing false-negatives caused by spelling errors and missing data while maintaining manageable computational complexity
Solution Approach 2:
The patent performs preliminary analysis of temporal patterns and their power spectra before the actual record linkage process. By pre-calculating and analyzing temporal pattern characteristics, the system prepares adaptive weights in advance, which then guide the matching process more effectively, reducing false-negatives without significantly increasing overall algorithm complexity
3Ease of operation
If Fellegi-Sunter method is used, then field-specific weights are uniform, but value-specific importance is not captured
Solution Approach 1:
The patent applies local quality by transitioning from uniform field-specific weights to value-specific weights that are locally adapted to the importance of individual values within temporal patterns. The power spectrum analysis identifies which temporal patterns are most significant for each specific value, allowing the system to assign different weights to different values based on their local importance, thereby improving value discrimination precision while maintaining operational simplicity
Data Source
AI summary
Methods, systems, and computer-readable media are provided for facilitating record matching and entity resolution and for enabling improvements in record linkage. A power-spectrum-based temporal pattern-specific weight may be incorporated into record linkage methods to enhance the record linkage accuracy and statistical performance. For example, in embodiments, a value-specific weight may be calculated from a population-based frequency of field-specific values and provides an opportunity to capture and measure the relative importance of specific values found in a field. A timeseries-derived Bayesian power spectrum weight may be calculated from the population-based frequency of temporal pattern-specific values in terms of intensities at various frequencies of the power spectrum computed from the timeseries, and further provides an opportunity to capture and measure the relative importance of specific sequences of care episodes.