Approximate Match Scoring for Biller Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic bill payment systems struggle with imperfect payee information, leading to failed electronic payments and the need for paper checks, which cause burdens and delays.
Innovation Solution
A computer system and method that approximate match payee information to biller information by determining an approximate match score, adjusting the match score based on account number comparisons, and generating a candidate list of billers for potential electronic payments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pure comparison of payee name to biller database is used, then matching accuracy is improved, but error tolerance deteriorates
Solution Approach 1:
The system changes the parameter of name comparison from exact matching to approximate matching by introducing a similarity scoring mechanism. This allows the system to tolerate variations in spelling, formatting, and minor errors while still identifying the correct biller through ranked candidate lists.
Solution Approach 2:
The matching system dynamically adjusts between exact matching and approximate matching based on the quality of input data. When exact matches are not found, the system transitions to approximate matching with similarity scoring, and can further adjust by incorporating account mask validation to refine results.
2Reliability
If approximate matching is implemented, then error tolerance is improved, but matching precision deteriorates
Solution Approach 1:
The system provides feedback through similarity scores and candidate rankings that indicate the confidence level of matches. This allows users to review and verify approximate matches, and the system can learn from corrections to improve future matching accuracy while maintaining high error tolerance.
Solution Approach 2:
The similarity scoring mechanism acts as an intermediary between exact matching and approximate matching. It provides a graduated approach where matches are ranked by confidence, allowing the system to maintain precision for high-confidence matches while tolerating errors in lower-confidence cases.
3Measurement precision
If multiple matching stages are implemented, then processing accuracy is improved, but system complexity deteriorates
Solution Approach 1:
The matching process is segmented into distinct stages: exact matching, approximate matching with similarity scoring, and account mask validation. Each stage handles specific types of matches independently, making the complex process more manageable and maintainable while improving overall accuracy.
Solution Approach 2:
The system performs preliminary exact matching before proceeding to more complex approximate matching. This preliminary action filters out obvious matches early, reducing the computational burden of subsequent stages and making the overall system more efficient despite its complexity.
4Reliability
If account mask comparison is added to adjust match scores, then matching reliability is improved, but processing time deteriorates
Solution Approach 1:
The system applies account mask comparison selectively rather than to all candidates. It first identifies promising candidates through name similarity scoring, then applies the more time-consuming account mask validation only to those candidates, achieving high reliability without processing all possible matches exhaustively.
Data Source
AI summary
A computer system including a processor, in communication with a database, programmed to identify a plurality of billers from the database, including biller information, and receive a payee file including payee information identifying a payee which issued a bill to a payor. The processor determines an approximate match score for each of the billers in a first set of billers of the plurality of billers by comparing payee information to biller information and identifies a candidate list of billers from the first set of billers having an approximate match score indicating a likelihood of matching the payee information. The processor adjusts the candidate list of billers by altering the match score of at least one candidate biller by comparing a payee account number of the payee information to an account mask associated with the candidate biller and generates an output file of candidate biller information from the adjusted candidate list.