Hierarchical Multi-Label Classification via Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods face challenges in efficiently processing large-scale multi-label documents without predefined hierarchies, leading to high computational complexity and reduced accuracy in classification tasks.
Innovation Solution
A method and system for hierarchical classification that automatically extracts hierarchies from documents using clustering processes, optimizing the hierarchy structure rather than the classifiers, and employing a top-down approach to decompose classification tasks into simpler sub-problems, utilizing tools like WEKA ML for improved performance measures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional flat classification methods are used for large-scale multi-label documents, then the classification can be performed with simple structure, but the computational complexity increases and accuracy decreases
Solution Approach 1:
The patent segments the flat classification problem into a hierarchical structure where the classification task is divided into multiple levels. Each level handles a subset of labels, breaking down the complex multi-label classification into simpler sub-problems. This segmentation reduces computational complexity at each level while maintaining overall classification accuracy through the hierarchical organization of label sets.
2Measurement precision
If hierarchical classification is implemented to improve accuracy, then classification performance improves, but the system complexity increases
Solution Approach 1:
The patent implements a dynamic hierarchy construction approach where the hierarchical structure is not fixed but adaptively created based on the specific classification task and data characteristics. The hierarchy can be adjusted and optimized for different multi-label datasets, allowing the system to balance accuracy improvement with manageable complexity by tailoring the hierarchical structure to each specific application scenario.
Data Source
AI summary
A technique of extracting hierarchies for multilabel classification. The technique can process a plurality of labels related to a plurality of documents, using a clustering process, to cluster the labels into plurality of clusterings representing a plurality of classes. The technique classifies the documents and predicts a plurality of performance characteristics, respectively, for the plurality of clusterings. The technique selects at least one of the clusterings using information from the performance characteristics and adds the selected clustering into a resulting hierarchy.