Graph Classification Model Training via Local and Global Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graph classification technologies limit user accessibility due to complex dataset formats, requiring coding skills and familiarity with machine learning backends, making it difficult for non-experts to prepare and understand datasets for training graph classification models.
Innovation Solution
A graphical user interface (GUI) is provided to guide users through selecting and importing datasets, analyzing configuration settings, and mapping categorical values to numeric targets, allowing users to select between local and global mapping schemes, and configuring settings for training graph classification models without manual coding or script editing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional machine learning backend configuration methods are used, then the system can process complex data formats, but the user accessibility and ease of operation deteriorate due to requiring coding skills and familiarity with backend terminology
Solution Approach 1:
The patent introduces an intermediary layer between the user and the machine learning backend. This intermediary translates user-friendly graphical interface selections into the complex configuration scripts and parameters required by the backend, allowing users to access complex data processing capabilities without needing to understand the underlying technical complexity
Solution Approach 2:
The patent creates simplified copies or representations of the complex backend configuration in the form of graphical user interface elements. Instead of requiring users to directly manipulate complex configuration files and scripts, the system presents simplified visual representations that mirror the essential functionality while hiding the underlying complexity
2Productivity
If detailed configuration scripts and text files are used, then the machine learning backend can be precisely controlled, but the time required for setup and configuration increases
Solution Approach 1:
The system performs preliminary actions by automatically generating and configuring the necessary scripts and parameters based on user selections in the graphical interface. Instead of requiring users to manually prepare detailed configuration files before model training, the system pre-processes these configurations in the background, significantly reducing the time users need to spend on setup
Solution Approach 2:
The system provides self-service functionality by automatically generating configuration scripts, selecting appropriate parameters, and preparing data formats based on user selections. The backend system serves itself by translating high-level user intentions into detailed technical configurations without requiring manual intervention, thereby accelerating the overall process
3Measurement precision
If specialized terminology and file formats are used, then the machine learning backend can maintain precision and control, but the difficulty of detecting and measuring understanding increases for domain experts unfamiliar with backend specifics
Solution Approach 1:
The system dynamically changes parameters based on user selections and data characteristics. Instead of requiring users to understand and manually configure all parameters in specialized file formats, the system automatically adjusts parameters like mapping schemes (local vs. global), data normalization options, and feature selection criteria based on the selected datasets and user preferences, maintaining precision while reducing user burden
Data Source
AI summary
A method may include directing display of a dataset menu listing datasets representative of graphs. The method may include identifying features in the datasets as corresponding to nodes and edges. The method may include selecting local or global mapping to map categorical feature values to numeric values. Local mapping may be selected in response to a distribution of feature values not corresponding across different graphs. Global mapping may be selected in response to a distribution of the feature values corresponding across different graphs. The method may include directing display of configuration settings that indicate the selection between local and global mapping for training a classification model. The method may include obtaining selected configuration settings. The method may include providing the selected configuration settings and datasets to a machine learning backend, which may utilize the machine learning algorithm, datasets, and selected configuration settings to train the classification model.