Linguistically Informed Graph Model and Semantic Contrastive Learning for Korean Short Text Classification

2026-03-03

AI Summary

This work addresses the challenges of Korean short-text classification, where performance is hindered by data sparsity, limited labeled resources, and the neglect of Korean’s agglutinative morphology and flexible word order in existing approaches. To overcome these limitations, the authors propose a novel heterogeneous graph neural network that integrates morphemes, part-of-speech tags, and named entities into a multi-level linguistic structure, explicitly embedding Korean-specific syntactic features into graph representations for the first time. Furthermore, they introduce SemCon, a semantic-aware contrastive learning framework, to refine decision boundaries. Extensive experiments on four Korean short-text datasets demonstrate that the proposed method significantly outperforms current state-of-the-art baselines, confirming the effectiveness of linguistically informed modeling and contrastive learning in enhancing text classification for agglutinative languages.

Technology Category

• Natural Language Processing: Lexical Semantics and Morphology
• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Short text classification (STC) remains a challenging task due to the scarcity of contextual information and labeled data. However, existing approaches have pre-dominantly focused on English because most benchmark datasets for the STC are primarily available in English. Consequently, existing methods seldom incorporate the linguistic and structural characteristics of Korean, such as its agglutinative morphology and flexible word order. To address these limitations, we propose LIGRAM, a hierarchical heterogeneous graph model for Korean short-text classification. The proposed model constructs sub-graphs at the morpheme, part-of-speech, and named-entity levels and hierarchically integrates them to compensate for the limited contextual information in short texts while precisely capturing the grammatical and semantic dependencies inherent in Korean. In addition, we apply Semantics-aware Contrastive Learning (SemCon) to reflect semantic similarity across documents, enabling the model to establish clearer decision boundaries even in short texts where class distinctions are often ambiguous. We evaluate LIGRAM on four Korean short-text datasets, where it consistently outperforms existing baseline models. These outcomes validate that integrating language-specific graph representations with SemCon provides an effective solution for short text classification in agglutinative languages such as Korean.

Problem

• Short text classification
• Korean language
• Agglutinative morphology
• Linguistic characteristics
• Contextual scarcity

Innovation

• hierarchical heterogeneous graph
• morpheme-level modeling
• agglutinative language
• semantic contrastive learning
• Korean short text classification