Account Number Validation via Density-Based Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic data entry validation systems for bank account numbers are inefficient in detecting invalid entries due to their reliance on prefix/suffix digit-based features and number length-based features, which fail to capture rare invalid account numbers, leading to user attrition and revenue loss.
Innovation Solution
The implementation of a machine learning (ML) technique that leverages the clustering nature of account numbers distributed by financial institutions, using algorithms like K-means clustering to identify valid and invalid ranges, thereby alerting users before submission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If prefix/suffix digit-based features and number length-based features are used for validation, then the validation system is simple to implement, but the accuracy of detecting invalid account numbers deteriorates
Solution Approach 1:
The patent transforms the validation approach from using simple digit-based features to using clustering-based parameters. By applying K-means clustering to group account numbers by their institutional sources, the system identifies valid ranges through density-based clustering, fundamentally changing the parameter space from individual digit features to collective range patterns.
Solution Approach 2:
The patent creates a simplified representation of valid account number patterns by generating synthetic valid account numbers based on identified clustering patterns. This copying approach allows the system to validate new entries by comparing them against learned patterns from training data, improving detection accuracy without requiring complex rule-based systems.
2Productivity
If traditional feature-based validation is used, then the system is computationally efficient, but the ability to capture rare invalid account numbers deteriorates
Solution Approach 1:
The patent implements a two-stage validation process where clustering patterns are pre-learned from training data before actual validation occurs. By performing K-means clustering and identifying valid ranges in advance, the system prepares lookup tables and pattern models that enable fast, reliable validation of new account numbers without requiring complex real-time analysis.
Solution Approach 2:
The patent replaces traditional mechanical rule-based validation with a machine learning-based pattern recognition system. By substituting deterministic digit-check rules with probabilistic clustering-based validation, the system achieves both high reliability in detecting rare invalid numbers and maintained computational efficiency through optimized pattern matching.
3Measurement precision
If density-based clustering is implemented, then the accuracy of identifying valid and invalid ranges is improved, but the computational complexity increases
Solution Approach 1:
The patent segments the account number space into distinct clusters based on their institutional sources using K-means clustering. By dividing the continuous space of possible account numbers into discrete, manageable clusters, the system identifies valid ranges for each institution separately, improving precision while making the overall problem more tractable through decomposition.
Solution Approach 2:
The patent develops a universal clustering framework that can handle multiple financial institutions with different account number formats and patterns. The K-means clustering algorithm serves multiple functions: grouping account numbers by institution, identifying valid ranges, and generating validation rules, thereby reducing overall system complexity through a single multi-functional approach.
Data Source
AI summary
An entry validation system executed by a processor, may compare an entry into a user interface (UI) field with at least one range of valid entries. The at least one range of valid entries may be generated by a machine learning (ML) system by clustering known valid data using a clustering algorithm producing a lowest number of clusters that absorbs all of the known valid data within the clusters. In response to the comparing, the entry validation system may provide a determination of validity or invalidity to an application displaying the UI field. The determination of validity may cause the entry to be processed by the application, and the determination of invalidity may cause the application to reject the entry.