Materialized View Reuse Through Semantic Query Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating materialized views are limited to syntactically identical expressions and are not scalable, requiring significant computational resources and time to execute queries against large data sets.
Innovation Solution
Utilizing machine learning-based techniques to identify semantically equivalent or containment relationships between query subexpressions, generating materialized views for these relationships, and returning pre-computed results to reduce query execution time and resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional syntactically identical expression matching is used to generate materialized views, then the system is simple to implement, but the scalability is poor and requires significant computational resources
Solution Approach 1:
The patent introduces machine learning models as an intermediary between query expressions and materialized view generation. The ML models learn semantic relationships and containment relationships from training data, enabling the system to identify applicable materialized views without requiring complex rule-based matching or exhaustive computation, thus improving scalability while managing system complexity through learned patterns
Solution Approach 2:
The patent transforms the matching criteria from strict syntactic identity to semantic equivalence and containment relationships. By changing the parameters of comparison from surface-level syntax to deeper semantic meaning, the system can generalize across different query formulations and identify materialized views that were previously inaccessible through traditional matching methods
2Loss of time
If materialized views are generated for more query expressions, then query execution time is reduced, but computational resources and time required for view generation increase
Solution Approach 1:
The patent performs preliminary action by pre-computing and storing materialized views for query expressions that are likely to be executed, using machine learning to predict which views will be beneficial. The system learns from historical query patterns and proactively generates materialized views before they are needed, reducing query execution time while avoiding waste of computational resources on views that would not be used
Solution Approach 2:
The patent implements feedback mechanisms where the system monitors query execution patterns and materialized view usage, using this information to continuously improve its ML models and refine which materialized views to generate. This feedback loop ensures that computational resources are invested in generating views that actually reduce query execution time, optimizing the trade-off between view generation cost and query performance
Data Source
AI summary
Embodiments described herein are directed to generating and returning materialized views for queries (or subexpressions thereof) having a particular relationship with each other. For instance, machine learning-based techniques may be utilized to identify query subexpressions that have at least one of a semantically equivalent relationship or a containment relationship with each other. Responsive to identifying such relationship(s), a materialized view may be generated for the identified subexpressions. When a query is subsequently received, machine learning-based techniques may be utilized to determine whether a subexpression of the query possesses at least one of a semantically equivalent relationship or a containment relationship with another subexpression for which a materialized view has been generated. Responsive to determining that such a subexpression of the query possesses one or more of such relationships, the materialized view generated for the other subexpression is returned.