Consolidating Data Type Repositories via Subtype Indicators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional frequent itemset mining techniques are inefficient in processing large datasets, leading to delayed results and increased computational time, which hampers the agility and competitiveness of IT companies in rapidly changing industries.
Innovation Solution
A system and method that utilize a model row generator to create subtype indicators and parent lists for XML data types, enabling the consolidation engine to combine and compare these lists to generate consolidation groups efficiently, thereby reducing redundant information and improving processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional frequent itemset mining techniques are used to process large datasets, then complete pattern extraction is achieved, but computational time increases and processing speed decreases
Solution Approach 1:
The patent applies preliminary action by pre-processing data into vertical formats with parent lists and subtype indicators before mining operations. Candidate itemsets are generated in advance based on parent list intersections, eliminating the need to explore all possible combinations during the mining phase. This preparatory structuring significantly reduces computational time while maintaining complete pattern extraction.
Solution Approach 2:
The patent segments the frequent itemset mining process into distinct phases: data pre-processing into vertical formats, candidate itemset generation based on parent list intersections, and final pattern mining. This segmentation allows each phase to be optimized independently, with candidate generation providing a filtered subset that reduces the overall computational burden while ensuring completeness.
2Measurement precision
If conventional horizontal or vertical frequent itemset mining approaches are used, then all patterns are extracted, but processing efficiency is reduced due to step-by-step candidate building
Solution Approach 1:
The patent performs preliminary action by pre-computing parent lists for each subtype indicator and generating candidate itemsets before the actual mining process. This allows the mining algorithm to work with a pre-filtered set of candidates rather than building them step-by-step during execution, significantly improving processing efficiency while maintaining complete pattern extraction.
Solution Approach 2:
The patent creates copied and structured representations of data in vertical formats with parent lists. Instead of repeatedly accessing and processing the original large dataset, the system works with these compact copied structures that contain only relevant information, thereby improving processing efficiency without losing any pattern extraction capability.
3Reliability
If large amounts of software artifacts are maintained to support product development, then software quality is improved, but development and maintenance costs increase
Solution Approach 1:
The patent merges redundant data type definitions by identifying and consolidating common parent types across multiple software artifacts. By combining shared structures into unified parent definitions, the system reduces the total volume of software artifacts while maintaining software quality through consistent reuse of consolidated components.
Solution Approach 2:
The patent creates universal parent data types that serve multiple functions across different software artifacts. These consolidated parent types can be reused by multiple child types and artifacts, reducing redundancy while ensuring software quality through consistent, standardized definitions that fulfill multiple purposes.
Data Source
AI summary
A method and system are described for generating consolidation groups. An example method includes obtaining a plurality of data type indicators of data types, wherein each data type indicator is associated with a hierarchical subtype of the data type; generating a first subtype indicator of a first hierarchical subtype associated with a first data type indicator in a first hierarchical subtype relationship; generating a first parent list including one or more representations of one or more of the data types associated with the first hierarchical subtype; generating a second subtype indicator of a second hierarchical subtype associated with a second data type indicator in a second hierarchical subtype relationship; generating a second parent list including a representation of one of the plurality of data types associated with the second hierarchical subtype; and generating a first consolidation group based on combining the first subtype indicator with the second subtype indicator.