Association Rule Mining for Supply Chain Data Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing supply chain management (SCM) software often reports irrelevant and overly complicated information, making it difficult for organizations to make effective changes to their supply chain and organizational operations.
Innovation Solution
The development of methods and systems that generate association rules for supplies within a supply database by determining the frequency of occurrence of unique combinations of use and product attributes, calculating a score based on confidence and lift, and ranking these combinations to provide actionable insights to SCM systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If SCM software reports detailed information about supply chain operations, then the quantity of information increases, but the relevance and simplicity of the information decreases
Solution Approach 1:
The patent segments the large volume of supply chain data into distinct association rules based on frequent itemset mining. Each rule represents a specific, relevant relationship between supply chain elements (e.g., products, suppliers, locations), transforming overwhelming raw data into discrete, actionable insights that maintain relevance while reducing information overload.
Solution Approach 2:
The system extracts only the most significant patterns and relationships from the supply chain data using confidence and lift metrics. By filtering out irrelevant associations and retaining only those that meet predetermined thresholds, the system separates signal from noise, providing concise information that directly addresses operational decision-making needs.
2Quantity of substance
If SCM software provides comprehensive supply chain data, then the quantity of information increases, but the complexity of the information structure increases
Solution Approach 1:
The patent divides complex supply chain data into structured association rules with clear components (antecedent, consequent, confidence, lift). This segmentation transforms an unwieldy information structure into organized, interpretable units that are easier to navigate and understand, reducing structural complexity while preserving comprehensive data insights.
Solution Approach 2:
The system transforms raw supply chain data into standardized parameters including confidence levels and lift values. By converting diverse data types into uniform metric-based rules, the system simplifies the information structure and enables consistent analysis across different supply chain dimensions without losing comprehensive coverage.
3Measurement precision
If association rules are generated with high confidence and lift scores, then the relevance of insights improves, but the computational processing time increases
Solution Approach 1:
The patent implements minimum threshold criteria for confidence and lift values, filtering out weak associations early in the analysis. By setting predetermined cutoff points (e.g., minimum confidence of 0.5, minimum lift of 1.0), the system focuses computational resources on generating only the most relevant high-precision rules, achieving accurate insights without exhaustive processing of all possible associations.
Solution Approach 2:
The system performs preliminary filtering of the supply chain data to identify frequent itemsets before generating association rules. By pre-processing the data to retain only frequently occurring combinations, the system reduces the search space for rule generation, thereby decreasing computational time while maintaining the precision of the final insights through subsequent confidence and lift calculations.
Data Source
AI summary
Provided are methods, systems, and apparatuses for improving supply chain management interfaces and functionality. A frequency of occurrence for each unique combination of a use attribute and a product attribute for various supplies may be determined. A frequency table containing each unique combination may be generated as well as one or more association rules. A level of confidence and a level of lift may be determined for each of the one or more association rules, both of which may be used to generate a score for each unique combination that may be used to rank each unique combination. The ranked unique combinations may then be provided to a supply chain management interface and a database.