AI Schema Matching for Semantic Database Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional schema matching methods rely solely on syntactic features, overlook semantic differences, and face scalability issues due to exhaustive matching, especially with dynamic data changes.
Innovation Solution
A two-stage schema matching approach using generative AI, involving table-level and attribute-level matching, with preliminary filtering and advanced data modeling to enhance accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional exhaustive schema matching is performed, then matching completeness is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent divides the schema matching process into two distinct stages: table-level matching (comparing table names, structures, and metadata) and attribute-level matching (comparing individual column attributes). This segmentation allows the system to perform comprehensive matching while reducing overall computational complexity by handling comparisons at appropriate granularities.
Solution Approach 2:
The patent performs preliminary filtering and table-level matching before conducting detailed attribute-level matching. By pre-identifying candidate table pairs based on structural similarity and metadata comparison, the system reduces the search space for subsequent attribute matching, thereby reducing computational complexity while maintaining matching completeness.
2Measurement precision
If traditional schema matching methods are used, then syntactic matching is achieved, but semantic differences are overlooked
Solution Approach 1:
The patent transforms the matching approach by incorporating multiple parameter types beyond simple syntax: data type semantics, nullability constraints, primary key/foreign key relationships, and cardinality rules. By evaluating multiple parameters with different weights and significance levels, the system achieves both syntactic precision and semantic understanding.
Solution Approach 2:
The patent creates a composite matching criterion that combines multiple factors: syntactic similarity (table names, column names), structural similarity (schema relationships), and semantic similarity (data types, constraints, meanings). This composite approach ensures both syntactic accuracy and semantic information retention.
3Reliability
If schema matching is performed on large-scale databases, then comprehensive coverage is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary table-level matching to identify candidate table pairs before conducting time-consuming attribute-level matching. By filtering out non-matching tables early based on structural and metadata comparisons, the system achieves comprehensive coverage while significantly reducing processing time for large-scale databases.
Solution Approach 2:
The patent segments the matching process into independent table-level and attribute-level stages that can be executed sequentially. This allows the system to process large databases efficiently by completing faster table-level comparisons first, then focusing computational resources on attribute-level matching only for candidate pairs.
4Stability of the object's composition
If dynamic data changes are handled through traditional methods, then data consistency is maintained, but adaptability to changes decreases
Solution Approach 1:
The patent implements a dynamic matching framework where the matching process can be re-executed when data changes occur. The system monitors schema modifications and triggers re-matching only for affected tables and attributes, maintaining data consistency while adapting efficiently to changes without requiring full re-matching of the entire database.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method for improved schema matching of two databases is disclosed. The method can receive a schema of a source table from a first database and a schema of a plurality of target tables from a second database, identify one or more matching tables among the plurality of target tables based on comparison of the schema of the source table and the schema of the plurality of target tables using a large language model, obtain first sample attribute data from the source table and second sample attribute data from a selected matching table, and identify one or more pairs of matching attributes between the source table and the selected matching table based on comparison of the first sample attribute data and the second sample attribute data using the large language model. Related systems and software for implementing the method are also disclosed.