Two-Stage Record Classification for Search Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computerized processes for searching and classifying large data sets, such as fingerprint databases, consume significant computing resources and time, necessitating the development of efficient methods to reduce computational overhead and search time.
Innovation Solution
The implementation of a two-stage search process, comprising a parsimonious search that filters records using a subset of features based on scoring metrics and a greedy search that examines the reduced subset with all features, to identify potential matches efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a comprehensive search using all features is performed on the entire data set, then matching accuracy is improved, but search time and computational resources increase significantly
Solution Approach 1:
The patent divides the search process into two distinct stages: a first stage using a subset of features for rapid filtering, and a second stage using all features for precise matching on a reduced set of candidates. This segmentation allows the system to achieve high matching accuracy while significantly reducing overall search time by avoiding comprehensive comparison of all records.
Solution Approach 2:
The first stage performs preliminary filtering using a subset of features to identify and eliminate non-matching records before the second stage applies all features. This preliminary action reduces the search space, allowing the computationally intensive full-feature comparison to be applied only to a small subset of potential matches, thereby maintaining accuracy while reducing time consumption.
2Measurement precision
If a comprehensive search using all features is performed on the entire data set, then matching accuracy is improved, but computational resources increase significantly
Solution Approach 1:
The patent segments the computational workload into two phases: an initial filtering phase using a subset of features that consumes minimal resources, and a detailed matching phase using all features applied only to filtered candidates. This segmentation dramatically reduces total computational resource usage while preserving matching accuracy.
Solution Approach 2:
The first stage performs preliminary filtering to eliminate clearly non-matching records before the resource-intensive second stage processes potential matches. This preliminary action prevents wasteful consumption of computational resources on records that would not yield positive results in a full comparison.
3Productivity
If a subset of features is used for searching, then search time and computational resources are reduced, but matching precision may deteriorate
Solution Approach 1:
The patent applies segmentation by using a subset of features in the first stage for efficient filtering, then transitioning to all features in the second stage for precise matching. This ensures that no precision is lost, as the full feature set is ultimately applied to the final candidate set.
Solution Approach 2:
The subset of features serves as a preliminary filtering mechanism rather than the final matching criterion. This preliminary action efficiently narrows the search space without compromising final precision, since all features are applied in the second stage to evaluate the reduced candidate set.
Data Source
AI summary
Techniques for classifying records in a dataset are described. The method includes identifying a set of features of a target record and assigning a scoring metric to each feature in the set of features. The method also includes processing the set of features based on the scoring metric to obtain a reduced subset of features and an expanded subset of features. The method also includes searching a store of electronic records using the reduced subset of features to obtain a reduced subset of electronic records. The method also includes searching the reduced subset of electronic records using the expanded subset of features to obtain a set of matching electronic records.