Interoperable Machine Learning Platform Schema Translation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine learning tools lack interoperability, resulting in incompatible data types and loss of metadata during translation between different systems, making it difficult to compose workflows using tools written in various programming languages within the same project.
Innovation Solution
An interoperable machine learning platform that enables seamless integration and execution of workflows across different programming languages like R, Python, and JAVA by using a bi-directional data and schema translation mechanism, allowing developers to compose and publish workflows as web services without manual translation coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If machine learning tools from different programming languages are integrated, then workflow composition capability is improved, but data type compatibility and metadata integrity deteriorate
Solution Approach 1:
The patent introduces a schema translation mechanism that acts as an intermediary between different programming language environments. This translation layer converts data types and metadata schemas between R, Python, Java and the platform's native format, preventing information loss while enabling workflow composition across languages.
Solution Approach 2:
The system dynamically changes data type parameters and metadata schema parameters during translation operations. By adjusting these parameters to match the target environment's requirements, the system maintains metadata integrity while enabling cross-language interoperability.
2Adaptability or versatility
If manual translation coding is required for interoperability, then programming language support is improved, but development complexity and time consumption worsen
Solution Approach 1:
The system provides self-service automation through its schema translation mechanism. When a workflow component from an external programming language is imported, the system automatically performs the translation and adaptation without requiring manual coding intervention, thereby supporting multiple languages while eliminating translation time consumption.
Solution Approach 2:
The platform pre-configures translation schemas and data type mappings for multiple programming languages (R, Python, Java). This preliminary preparation enables automatic translation without requiring users to write translation code, thus improving language support while preventing time loss.
3Adaptability or versatility
If data formats are converted between systems, then system compatibility is improved, but information loss increases
Solution Approach 1:
The schema translation mechanism serves as an intermediary that preserves data information during format conversion. It maintains the semantic meaning and structure of data while adapting it to different system requirements, thereby achieving system compatibility without information loss.
Data Source
AI summary
An interoperable platform that provides a way to automatically compose and execute even complex workflows without writing code is described. A set of pre-built functional building blocks can be provided. The building blocks perform data transformation and machine learning functions. The functional blocks have few well known plug types. The building blocks can be composed to build complex compositions. Interoperability between data formats, metadata schema and interfaces to machine learning (ML) functions and trained machine learning models can be provided with no loss of information. A cloud runtime environment can be provided in which the composed workflows can be hosted as REST API to run in production.