Series Data Likelihood Ratios for Identity Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems for identity verification and classification lack efficiency in determining whether series data are derived from the same target as predetermined data, particularly in handling biometric information.
Innovation Solution
An information processing apparatus and method that calculates a likelihood ratio based on the degree of similarity or difference between series data and predetermined data, using elements from a plurality of consecutive data points, and determines the target origin of the series data based on this ratio.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If likelihood ratio is calculated using only biometric information matching scores, then the verification process is simple, but the accuracy of determining target origin is insufficient
Solution Approach 1:
The patent combines multiple types of data (biometric information matching scores, series data, and predetermined data) into a unified likelihood ratio calculation framework. By merging these data sources and calculating their joint probability ratios, the system achieves more accurate target origin determination while maintaining a structured calculation approach that manages complexity through systematic integration of multiple data types.
Solution Approach 2:
The patent transforms the verification process by changing the parameters used in likelihood ratio calculation from simple matching scores to a comprehensive set of parameters including series data characteristics and predetermined data comparisons. This parameter expansion enables the system to capture more nuanced patterns in biometric data, improving measurement precision without overwhelming complexity.
2Reliability
If series data are analyzed using only single-point comparisons, then the processing speed is high, but the reliability of identity verification is reduced
Solution Approach 1:
The patent segments the series data into multiple consecutive elements and analyzes them individually within the likelihood ratio calculation. By dividing the data stream into discrete segments that can be processed independently, the system maintains high processing speed while achieving reliable verification through the cumulative analysis of multiple segments. Each segment contributes to the overall likelihood ratio, enabling parallel processing opportunities.
Solution Approach 2:
The patent implements continuous analysis of series data by maintaining a sliding window that processes consecutive elements in sequence. This continuous action allows the system to analyze multiple data points without resetting or restarting processing, thereby maintaining high productivity while improving reliability through comprehensive data coverage. The continuous likelihood ratio calculation integrates all segments smoothly, avoiding processing interruptions.
3Measurement precision
If the system considers degree of similarity or difference between series data and predetermined data, then the classification accuracy improves, but the calculation complexity increases
Solution Approach 1:
The patent introduces the likelihood ratio as an intermediary calculation that mediates between the series data and predetermined data. Instead of directly comparing and analyzing similarity and difference complexities, the system uses the likelihood ratio formula as an intermediary step that encapsulates these comparisons in a standardized manner. This intermediary approach simplifies the overall calculation structure while maintaining high classification accuracy through the systematic evaluation of data relationships.
Data Source
AI summary
An information processing apparatus includes: an acquisition unit that obtains a plurality of elements included in series data; a calculation unit that calculates a likelihood ratio indicating a likelihood that the series data are derived from the same target as that of predetermined data, on the basis of at least two consecutive elements of the plurality of elements; and a determination unit that determines whether or not the series data are derived from the same target as that of the predetermined data, on the basis of the likelihood ratio. The calculation unit calculates the likelihood ratio in view of a degree of similarity or difference between the series data and the predetermined data. According to the information processing apparatus, it is possible to accurately determine whether or not the series data is derived from the same target as that of the predetermined data.