Machine Learning Matching of Post-Approved and Pre-Approved Transactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large entities face challenges in identifying procurable expenses from transaction records due to the computational expense and memory requirements of comparing large numbers of transactions, which can lead to inefficiencies and increased costs when goods or services are externally procured instead of through approved channels.
Innovation Solution
A system utilizing machine learning algorithms to normalize and match post-approved transaction records with pre-approved records, reducing the number of operations required by the server while maintaining high accuracy, by categorizing transactions and using clustering algorithms to efficiently compare records across entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If transaction-by-transaction comparison is performed to identify procurable expenses, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The patent segments the large set of transactions into smaller batches or groups that can be processed in parallel or sequentially with reduced memory requirements. By dividing the comparison task into manageable segments, the system achieves thorough matching accuracy while controlling computational resource consumption.
Solution Approach 2:
The system performs preliminary filtering and preprocessing of transactions before the main comparison operation. By pre-categorizing, pre-matching, or pre-filtering transactions based on obvious criteria, the system reduces the scope of detailed comparisons needed, thereby maintaining accuracy while reducing overall computational energy expenditure.
2Measurement precision
If more transaction records are compared to identify procurable expenses, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent implements periodic or iterative processing where transactions are compared in cycles or batches rather than all at once. This allows the system to process comprehensive transaction sets while managing time constraints through structured periodic evaluation and early termination when confidence thresholds are met.
Solution Approach 2:
The system performs partial comparisons focused on the most promising or high-value transactions first, rather than exhaustively comparing all possible pairs. By applying excessive action to critical subsets and minimal action to lower-priority comparisons, the system achieves sufficient precision within acceptable timeframes.
3Measurement precision
If transaction records are stored and compared across multiple entities, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal matching framework that handles multiple entities, transaction types, and comparison criteria through a single integrated system. This multi-functional approach achieves cross-entity matching precision while avoiding the complexity of separate specialized systems for each entity or transaction category.
Solution Approach 2:
The system dynamically adjusts comparison parameters, thresholds, and matching criteria based on the specific entity, transaction type, or context being analyzed. By changing parameters adaptively rather than using fixed complex rules for all cases, the system achieves high precision across diverse scenarios while maintaining manageable system complexity.
Data Source
AI summary
Systems and methods for improving a computing system comparing past post-approved transaction records to past pre-approved transaction records are described herein. In an embodiment, a server stores a first plurality of digital electronic records identifying a plurality of past post-approved transactions by a first entity and a second plurality of digital electronic records for the first entity identifying a plurality of past pre-approved transactions by the first entity. The server uses a first machine learning system to determine that a subset of the first plurality of digital electronic records which identify a subset of the plurality of past post-approved transactions correspond to one or more pre-identified categories and a second machine learning system to match one or more particular digital electronic records of the subset of the first plurality of digital electronic records with one or more records of the second plurality of digital electronic records.