Relational Database Model Management via Dynamic Table Creation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in efficiently storing and managing a large number of structurally diverse predictive models in relational databases, leading to excessive storage space, CPU, and memory consumption, particularly in virtual drug screening and other model-based prediction scenarios.
Innovation Solution
A method and system that dynamically create database tables in a relational database based on specific model requirements, allowing models to be stored in a non-structured format, with only the best-model stored persistently, and other models created temporarily as needed, optimizing storage and CPU usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all models are stored persistently in the database, then model availability and reliability are improved, but storage space and memory consumption increase significantly
Solution Approach 1:
The patent extracts only the essential model data (best-model) from the complete model set and stores it persistently in the database. Other models are loaded temporarily from external sources only when needed for training or evaluation, rather than storing all models in the database simultaneously. This selective extraction resolves the contradiction by maintaining availability of the best model while avoiding excessive storage consumption.
Solution Approach 2:
The patent employs temporary model storage mechanisms where non-best models are loaded into memory or temporary storage only when required for specific training or evaluation tasks, then discarded afterward. This disposable approach allows flexible model access without committing permanent storage resources for all models, thus resolving the storage availability contradiction.
2Adaptability or versatility
If database tables are created for each model, then model management flexibility is improved, but device complexity and CPU overhead increase
Solution Approach 1:
The patent implements a universal database table structure that can accommodate multiple models without requiring separate specialized tables for each model. The single table design with appropriate schema handles diverse model types and structures, providing flexibility through a standardized interface rather than through structural complexity. This resolves the contradiction by achieving adaptability without increasing device complexity.
3Speed
If all models are loaded into memory, then model access speed is improved, but memory consumption increases excessively
Solution Approach 1:
The patent applies local quality by loading only the specific model data needed for current operations into memory, rather than loading all models simultaneously. The best-model is kept in memory for rapid access, while other models are loaded on-demand from external storage when required for training or evaluation. This selective loading approach maintains fast access speeds for active models while keeping overall memory consumption manageable.
Data Source
AI summary
A system includes a model repository comprising a plurality of models respectively being adapted to perform, when used by an analytical program, a computational task, in which a first database table is created in the database, the first database table having a predefined table structure that corresponds to the analytical program, a best-model of the plurality of models is stored in the first database table, and a request of a client device to perform the computational task and comprising input data is received. If the received request does not comprise a model-ID, the analytical program reads the model currently stored in the first table and uses the read model for performing the computational task on the input data. If the received request comprises a model-ID, the analytical program creates a second database table having the predefined table structure in the database, reads a model associated with the model-ID from the model repository, stores the read model associated with the model-ID in the second table, and uses the model read from the second table for performing the computational task on the input data.