ML Pipeline for Hierarchical Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The variability in terminology across different entities and over time complicates data analysis, leading to uncertain meaning and inefficient query results, as existing approaches fail to apply consistent terminology effectively.
Innovation Solution
A system utilizing a trained machine learning pipeline with multiple independently trained models to assign hierarchical classifications to data items, including a cluster-based and classification-based model, and a terminology normalizer to convert inconsistent terminology to standard terms, facilitating accurate and consistent analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple independently trained machine learning models are used to assign hierarchical classifications, then analytical accuracy is improved, but system complexity increases
Solution Approach 1:
The system segments the classification task into multiple independent machine learning models, each trained separately to perform specific classification functions. This segmentation allows each model to specialize in particular aspects of data classification, improving overall accuracy while maintaining manageable complexity through modular design.
Solution Approach 2:
The patent implements a hierarchical classification structure where classifications are organized in nested levels. Multiple models operate at different hierarchical levels, with each model's output feeding into the next level. This nesting approach improves accuracy by capturing complex patterns at multiple granularities while organizing system complexity in a structured manner.
2Stability of the object's composition
If terminology normalization is applied to convert inconsistent terminology to standard terms, then data analysis consistency is improved, but processing time increases
Solution Approach 1:
The system performs terminology normalization as a preliminary action before the main classification and analysis processes. By pre-converting inconsistent terminology to standard terms, the system ensures consistency throughout subsequent processing steps without repeatedly performing normalization operations, thereby reducing total processing time while maintaining terminology stability.
Data Source
AI summary
Techniques are disclosed for using a trained machine learning (ML) pipeline to identify categories associated with target data items even though the identified categories may not already be present in the hierarchy. The ML pipeline may include trained cluster-based and classification-based machine learning models, among others. If the results of the cluster-based and classification-based machine learning models are the same, then the target data items is assigned to a hierarchical classification consistent with the identical results of the machine learning model. An assigned hierarchical classification may be validated by the operation of subsequent trained ML models that determine whether parent and child categories in the identified classification are properly associated with one another.