Predicted Similarity Score Generation for Input Sequences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semantic relatedness processing systems are inefficient and unreliable in determining similarity between input sequence pairs, which hinders effective database integration and consolidation.
Innovation Solution
A method is introduced to dynamically generate a predicted similarity score for input sequence pairs by calculating token-level similarity probability scores, fuzzy match scores, character-level match scores, and harmonic mean scores, using machine learning models and neural network representations to determine the similarity between sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing semantic relatedness processing systems are used to determine similarity between input sequence pairs, then the processing can be performed, but the system is inefficient and unreliable
Solution Approach 1:
The patent segments the similarity determination process into multiple independent scoring components: token-level similarity probability score, target region match indication, fuzzy match score, character-level match score, and harmonic mean score. Each component processes specific aspects of sequence comparison independently, then combines results to produce the final predicted similarity score. This segmentation improves both reliability through comprehensive analysis and efficiency through parallel processing of discrete components.
2Reliability
If multiple scoring mechanisms are implemented to improve similarity determination accuracy, then reliability improves, but system complexity increases
Solution Approach 1:
The patent merges five distinct scoring mechanisms into a unified predicted similarity score through a machine learning model. The token-level similarity probability score, target region match indication, fuzzy match score, character-level match score, and harmonic mean score are combined using weighted aggregation and neural network processing. This merging approach maintains high reliability by incorporating multiple analysis dimensions while managing complexity through a standardized integration framework.
Solution Approach 2:
The patent creates a universal similarity determination system that handles multiple types of sequence comparisons through a single multi-functional model. The system processes token-level semantics, region matching, fuzzy string matching, and character-level analysis within one framework, making it adaptable to various input sequence types and comparison scenarios without requiring separate specialized systems for each function.
3Measurement precision
If comprehensive semantic analysis is performed to improve similarity accuracy, then measurement precision improves, but computational cost increases
Solution Approach 1:
The patent implements partial action by selectively applying different scoring mechanisms based on the specific comparison needs. The system can adjust the depth of analysis - using only essential scoring components for routine comparisons while engaging full comprehensive analysis only when higher precision is required. This approach maintains measurement precision when needed while reducing computational energy consumption for standard operations.
Data Source
AI summary
Systems and methods for dynamically generating a predicted similarity score for a pair of input sequences. A predicted similarity score for a pair of input sequences is determined based at least in part on at least one of a token-level similarity probability score for the pair of input sequences, a target region match indication for the pair of input sequences, a fuzzy match score for the pair of input sequences, a character-level match score for the pair of input sequences, one or more similarity ratio occurrence indicators for the pair of input sequences, and a harmonic mean score of the fuzzy match score for the pair of input sequences and the token-level similarity probability score for the pair of input sequences.