Core Concepts
Hierarchical text classification with minimal supervision using Taxonomy Enrichment and LLM enhancement.
Abstract
Hierarchical text classification is essential in text mining.
TELEClass enhances label taxonomy with class-indicative terms.
LLMs are used for data annotation and tailored for hierarchical label space.
Experiments show TELEClass outperforms previous methods.
Stats
Hierarchical text classification aims to categorize each document into a set of classes in a label taxonomy.
Most earlier works focus on fully or semi-supervised methods that require a large amount of human annotated data.
TELEClass can outperform previous weakly-supervised hierarchical text classification methods and LLM-based zero-shot prompting methods on two public datasets.
Quotes
"Most earlier works tackle this task in fully supervised or semi-supervised settings."
"Experiments show that TELEClass can outperform previous weakly-supervised hierarchical text classification methods."