Fuzzy String Alignment via Term Distance Cost Matrix
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computing systems face challenges in accurately and efficiently applying fuzzy logic due to their binary nature and limited resources, particularly in tasks like string matching, where semantics and meaning are not adequately considered by general rule-based approaches.
Innovation Solution
A method and system for computing term distances between string terms, generating a cost matrix, selecting candidate alignments, and outputting a match identifier, which includes tokenization, term distance calculation, cost matrix generation, alignment scoring, and selection, allowing for flexible algorithms based on resource usage and accuracy requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If general rule-based approaches are used for string matching, then the system is simple and fast, but it does not consider semantics and meaning of strings
Solution Approach 1:
The patent segments strings into individual terms and compares them pairwise, computing term distances for each term pair. This segmentation allows the system to consider semantic meaning of individual terms while maintaining computational efficiency through modular processing of term pairs rather than treating entire strings as single units.
Solution Approach 2:
The patent introduces a cost matrix as an intermediary structure that stores term distances between all pairs of terms from two strings. This cost matrix serves as a mediator between the raw string data and the final alignment decision, enabling semantic comparison through structured intermediate representation that captures term-level relationships.
2Measurement precision
If fuzzy logic is applied with high accuracy, then string matching precision improves, but computing resource consumption increases
Solution Approach 1:
The patent generates multiple candidate alignments and scores them, but only processes a limited number of top candidates rather than exhaustively evaluating all possible alignments. This partial action approach provides high accuracy by considering multiple possibilities while controlling computational resources by limiting the number of candidates processed to a manageable subset.
Solution Approach 2:
The patent computes term distances using configurable parameters such as keyboard adjacency thresholds and vowel matching rules, allowing the system to adjust the level of fuzzy matching based on resource availability and accuracy requirements. These parameter changes enable flexible control over computational intensity versus matching precision.
3Measurement precision
If multiple candidate alignments are generated and scored, then matching accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary computation of term distances between all term pairs and stores them in a cost matrix before generating candidate alignments. This preliminary action pre-processes the data in a way that speeds up subsequent alignment generation and scoring, reducing the time penalty of evaluating multiple candidates by having term distances readily available.
Solution Approach 2:
The patent replaces exhaustive mechanical enumeration of all possible alignments with a scoring model that efficiently evaluates candidate alignments based on pre-computed term distances. This substitution uses a mathematical scoring approach instead of brute-force mechanical comparison, reducing processing time while maintaining accuracy through intelligent candidate selection and scoring.
Data Source
AI summary
A method includes computing multiple term distances between pairs of multiple first string terms in a first string and multiple second string terms in a second string, generating a cost matrix based on the term distances, and selecting a set of candidate alignments based on the cost matrix. The method further includes generating multiple alignment scores for the set of candidate alignments, and selecting, from the set of candidate alignments, an alignment between the first string and the second string based on the alignment scores. The method further includes outputting a match identifier based on the alignment.