MMD-Sense-Analysis: Word Sense Detection Leveraging Maximum Mean Discrepancy

2025-06-02

AI Summary

This study addresses lexical semantic change detection, aiming to automatically identify and interpretably characterize semantic shifts of words across historical corpora. Existing methods lack both quantitative measures of cross-temporal semantic divergence and the capacity to identify salient dimensions underlying such shifts. To bridge this gap, we introduce Maximum Mean Discrepancy (MMD) into lexical change detection for the first time, integrating temporal semantic vector modeling with kernel-based interpretable variable selection to construct an end-to-end framework for quantifying and attributing semantic drift. Experiments demonstrate that our approach significantly outperforms baselines in both change-word identification accuracy and interpretability of evolutionary trajectories. Crucially, it precisely pinpoints the core semantic dimensions driving lexical change. The framework thus provides a novel tool for historical semantics and computational linguistics—one that combines statistical rigor with cognitive transparency.

Technology Category

• Natural Language Processing: Lexical Semantics and Morphology
• Data Mining & Knowledge Management: Semantic Web
• Machine Learning: Dimensionality Reduction/Feature Selection

Application Category

• Web Mining and Content Analysis: Models for Web evolution
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Word sense analysis is an essential analysis work for interpreting the linguistic and social backgrounds. The word sense change detection is a task of identifying and interpreting shifts in word meanings over time. This paper proposes MMD-Sense-Analysis, a novel approach that leverages Maximum Mean Discrepancy (MMD) to select semantically meaningful variables and quantify changes across time periods. This method enables both the identification of words undergoing sense shifts and the explanation of their evolution over multiple historical periods. To my knowledge, this is the first application of MMD to word sense change detection. Empirical assessment results demonstrate the effectiveness of the proposed approach.

Problem

• Detects word sense changes over time
• Uses Maximum Mean Discrepancy for semantic analysis
• Explains word meaning evolution across historical periods

Innovation

• Uses Maximum Mean Discrepancy for sense detection
• Quantifies word meaning changes over time
• First MMD application in sense shift analysis