Specialized Serializer Type Inference for Big Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database management systems face difficulties in handling big data due to excessive processing loads caused by dynamic serialization and deserialization processes, which are time-consuming and costly, especially when dealing with semi-structured data like JSON.
Innovation Solution
A method and apparatus for performing type inference of serialization for each generation site, specializing serializers, and generating dedicated serializers or deserializers to reduce processing time by statically inferring types during query compilation, rather than dynamically checking types and values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If dynamic serialization and deserialization are executed while checking types and values, then data can be processed with flexibility, but processing time and computational cost increase excessively
Solution Approach 1:
The patent applies preliminary action by performing type inference during the query compilation phase rather than during data serialization/deserialization execution. The system infers and determines types in advance, storing this type information in a type information management unit. When serialization occurs, the pre-determined types are directly used without runtime type checking, thus eliminating the time-consuming dynamic type verification while maintaining processing flexibility.
2Reliability
If type checking is performed dynamically during serialization, then data integrity is maintained, but computational cost and processing overhead increase
Solution Approach 1:
The system performs type inference and validation in advance during query compilation, storing the determined types in the type information management unit. This preliminary type determination ensures data integrity is verified before actual data processing occurs, eliminating the need for costly runtime type checking while maintaining reliability.
Solution Approach 2:
The patent replaces the mechanical runtime type checking process with a compilation-time type inference system. Instead of dynamically checking types during serialization/deserialization operations, the system uses static type information determined during query compilation, substituting runtime verification with pre-computed type data.
3Device complexity
If a generalized serializer is used for all generation sites, then device complexity is reduced, but processing time increases due to excessive type checking
Solution Approach 1:
The patent applies local quality by creating specialized serializers tailored to each generation site's specific type requirements. The type information management unit stores type information specific to each generation site, allowing the system to generate and use customized serializers that match the local data characteristics. This avoids the need for complex runtime type checking while maintaining low device complexity through targeted specialization.
Data Source
AI summary
To provide a method, apparatus, and computer program for performing type inference of serialization for each generation site and specializing a serializer for each generation site. A type of serialization is inferred for each generation site of compiling a query, and a serializer is specialized for each generation site based on the inferred type and a type that is actually used. A data value is serialized using the specialized serializer for each generation site. The inference is executed while transcribing identification information assigned to each generation site to a type as an annotation, and the inferred type and the type used in serialization are recursively compared.