Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering

2026-07-11

AI Summary

This work addresses a critical limitation in existing optimal transport–based short text clustering methods, which often neglect semantic consistency among samples, leading to semantically similar instances being assigned disparate pseudo-labels and thereby degrading clustering performance. To overcome this issue, the paper proposes a novel clustering framework that jointly models local semantic consistency and global cluster structure for the first time. Specifically, an instance-level attention mechanism is introduced to capture neighborhood semantic relationships, which are then integrated into the optimal transport process to generate pseudo-labels that balance both local coherence and global structural integrity. Coupled with pseudo-label–guided self-supervised learning, the proposed method achieves significant performance gains over state-of-the-art approaches across multiple benchmark datasets.

Abstract

Pseudo-labeling based on Optimal Transport (OT) has become an effective mechanism for enhancing short text clustering. Existing OT methods are short in modeling semantic consistencies between samples, which may assign different pseudo-labels to semantically similar samples. These erroneous pseudo-labels can cause the model to produce inferior clusters. This paper proposes a novel short text clustering framework, which remedies the neglect of semantic consistency in existing OT methods, generating reliable pseudo-labels to facilitate clustering. Specifically, the proposed approach first designs an instance-level attention mechanism to capture semantic relationships between samples, which are then integrated into the OT formulation to endow the transport process with neighborhood semantic awareness. By solving the proposed OT formulation, reliable pseudo-labels are obtained that simultaneously account for sample-to-sample semantic consistency and sample-to-cluster global structure information. These pseudo-labels are then used as supervisory signals to guide the model to achieve accurate clustering. Extensive experiments demonstrate that the proposed approach outperforms state-of-the-art methods. The code is available at: \href{https://github.com/YZH0905/CAOT-STC}{https://github.com/YZH0905/CAOT-STC}.

Problem

• short text clustering
• optimal transport
• pseudo-labeling
• semantic consistency
• global structure

Innovation

• Optimal Transport
• Short Text Clustering
• Semantic Consistency
• Instance-level Attention
• Pseudo-labeling