Machine Learning Clustering for Expense Record Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in efficiently identifying and matching post-approved transaction records with pre-approved records across large databases, leading to increased computational costs and memory requirements, particularly in managing travel expenses and contracts, where many potential procurement opportunities go unnoticed.
Innovation Solution
A server-based system utilizing machine learning algorithms to normalize and match post-approved transaction records with pre-approved records, both within an entity and across entities, reducing the number of comparisons needed by categorizing and clustering data, thereby minimizing processing power and time while identifying procurable expenses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If transaction-by-transaction comparison is performed to identify procurable expenses, then identification accuracy is improved, but computational cost and memory requirements increase significantly
Solution Approach 1:
The patent segments the large database of pre-approved transactions into smaller clusters using machine learning algorithms. Instead of comparing each post-approved transaction against all pre-approved transactions individually, the system first groups pre-approved transactions into clusters based on similarity, then only compares against relevant clusters. This segmentation dramatically reduces the number of comparisons needed while maintaining identification accuracy.
Solution Approach 2:
The system performs preliminary actions by pre-processing and clustering the pre-approved transaction data before the actual matching process. Machine learning models are trained in advance to understand transaction patterns, and the pre-approved transactions are organized into clusters beforehand. This preliminary preparation enables faster, more efficient matching without sacrificing accuracy during the actual expense identification process.
2Productivity
If more transaction records are compared to identify procurable expenses, then the number of identified opportunities increases, but processing time and computational resources increase
Solution Approach 1:
By segmenting the transaction database into meaningful clusters based on transaction characteristics, the system can efficiently process large numbers of records without linearly increasing processing time. The segmentation allows parallel processing and reduces the search space for each comparison, enabling the system to identify more opportunities while maintaining acceptable processing speeds.
Solution Approach 2:
The system changes parameters by using machine learning to dynamically adjust comparison thresholds and cluster groupings based on the specific characteristics of the transactions being analyzed. This allows the system to optimize the balance between identifying comprehensive opportunities and maintaining efficient processing speeds, adapting to different data sets and scenarios.
3Reliability
If comprehensive transaction comparison is performed across all entities, then procurement opportunities are better identified, but memory requirements and computational complexity increase
Solution Approach 1:
The patent applies segmentation by dividing the comprehensive transaction data across multiple entities into entity-specific clusters and then performing inter-entity cluster comparisons. This approach maintains the ability to identify cross-entity procurement opportunities while reducing the immediate computational burden on any single system, as the complexity is distributed and managed through structured clustering rather than brute-force comparison.
Data Source
AI summary
In one embodiment, a method can access an expense line for a travel expense for an enterprise. The expense line is the subject of a procurement action under a relevant contract. The method can use a first machine learning model, determining a category associated with the expense line. The method can evaluate the expense line using the category associated with the expense line and a table of categories that are procurable with references to related travel contracts. The method can determine a match between a description field of the expense line and description fields of historic invoice records associated with travel contracts. In response to determining the match, the method can execute a responsive action for the expense line, wherein the responsive action includes generating a notification to a user, marking the expense line, and initiating a workflow to apply the contract to the expense line.