US 12,405,779 B2
Ontology-based multi-programming language component specifications and workflow system and use method
Sizhu Wu, Beijing (CN); Qing Qian, Beijing (CN); Shengyu Liu, Beijing (CN); Xiaolei Xiu, Beijing (CN); and Anran Wang, Beijing (CN)
Assigned to Institute of Medical Information, Chinese Academy of Medical Sciences, Beijing (CN)
Filed by Institute of Medical Information, Chinese Academy of Medical Sciences, Beijing (CN)
Filed on Apr. 21, 2024, as Appl. No. 18/641,383.
Claims priority of application No. 202311499735.9 (CN), filed on Nov. 13, 2023.
Prior Publication US 2025/0156165 A1, May 15, 2025
Int. Cl. G06F 8/51 (2018.01); G06F 8/41 (2018.01)
CPC G06F 8/51 (2013.01) [G06F 8/41 (2013.01)] 8 Claims
OG exemplary drawing
 
1. An ontology-based multi-programming language component specifications and workflow system, comprising a central processing unit (CPU), a graphics processing unit (GPU), a memory, a hard disk, a multi-language programming component manager, a multi-language programming component workflow design modeler, a multi-programming language transformation engine and a core workflow engine, wherein the multi-language programming component manager is connected with the multi-language programming component workflow design modeler and the multi-programming language transformation engine; and the core workflow engine is connected with the multi-language programming component workflow design modeler and the multi-programming language transformation engine;
the multi-language programming component manager encapsulates, presets, organizes, manages and stores multi-type and multi-programming language components in a platform; wherein the multi-type and multi-programming language components are codes in different programming languages to implement specific functions;
the multi-language programming component workflow design modeler supports design and construction of a workflow model containing a mixture of the components in different programming languages;
wherein the different programming languages comprises Java language, Python language, and R language; wherein data types in Java language comprises int, long, double, Boolean, char, string, and list; wherein data types in Python language comprises numbers, Boolean, string, and list; wherein data types in R language comprises double, integer, Boolean, char, list;
when performing data communication across components in different programming languages, the multi-programming language transformation engine converts and maps the data types and parameters required by the components of any programming language of the different programming languages to the data types and parameters of another programming language of the different programming languages according to requirements; and
the core workflow engine coordinates multiple node tasks of different programming language components and their data dependencies,
wherein in the multi-language programming component workflow design modeler, node types available for workflow modeling are defined as follows:
start node: can only be used at a beginning of the workflow, any nodes in front of the start node are not allowed, but nodes can be connected after the start node;
end node: can only be used at an end of the workflow, nodes in front of the end note are allowed, but any nodes after the end node are not allowed;
merging node: an entry is connected to at least two nodes, and an exit is connected to only one node;
diverging node: only one node entry is provided, but at least two nodes are provided as node exits;
diverging and merging node: at least two node entries and node exits are respectively provided; and
common node: only one entry node and one exit node are respectively provided,
wherein, in the system, data reading components and data preprocessing analysis components are constructed with the Java language, wherein data reading is a fixed start node; machine learning classification algorithm components are constructed with the Python language, wherein the machine learning classification algorithm components are all common nodes or diverging and merging nodes; and data visualization analysis components are constructed with the R language, wherein the data visualization analysis components are all end nodes.