Procurement Request Classification Using Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current procurement processes are inefficient and often result in delays due to their manual nature and the lack of expertise among individuals handling procurement requests, leading to an undesirable user experience and increased risk of non-compliance.
Innovation Solution
A system and method that utilize a trained classifier to determine the complexity of procurement requests and assign them to appropriate procurement professionals based on workload, incorporating machine learning algorithms like decision trees, random forests, and XGboost classifiers to ensure accurate matching and efficient processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a manual procurement process is used, then the process is simple to implement, but it results in delays and lacks expertise in handling requests
Solution Approach 1:
The patent replaces the manual mechanical procurement process with an automated machine learning-based system. The classifier model automatically evaluates procurement requests and assigns them to appropriate professionals, eliminating manual evaluation delays and ensuring expert handling while maintaining system manageability through structured automation.
Solution Approach 2:
The system enables self-service through automated classification and assignment. The machine learning model independently evaluates request complexity and routing decisions without human intervention in the initial assessment phase, allowing the system to serve itself in the decision-making process while improving overall efficiency.
2Reliability
If procurement requests are handled by individuals without expertise, then the process is easier to operate, but it increases the risk of non-compliance
Solution Approach 1:
The machine learning classifier acts as an intermediary between the procurement request and the handling professional. It mediates the assignment process by evaluating request characteristics and matching them with appropriately skilled professionals, ensuring compliance expertise is automatically applied without complicating the overall operational flow.
Solution Approach 2:
The system performs preliminary classification and expertise matching before the procurement request reaches the handling professional. By pre-evaluating request complexity and assigning to qualified individuals in advance, the system ensures compliance accuracy is established beforehand rather than being checked during the process.
3Measurement precision
If a complex classification system is used to assign procurement requests, then the accuracy of professional matching improves, but the system complexity increases
Solution Approach 1:
The patent uses a trained machine learning classifier model that copies and generalizes patterns from historical procurement data. The model learns from past successful assignments and applies these patterns to new requests, achieving high measurement precision through learned patterns rather than complex manual evaluation rules.
Solution Approach 2:
The system transforms complex procurement requests into measurable parameters and features that the machine learning model can process. By converting qualitative request characteristics into quantifiable parameters for classification, the system achieves accurate matching while keeping the underlying model structure manageable and interpretable.
Data Source
AI summary
A procurement system allows a user to provide a request for goods or services. The request is processed to determine its complexity and, for high complexity cases, select an appropriate procurement professional using a trained classifier to handle the procurement request.