Improving Interpretability of Lexical Semantic Change with Neurobiological Features

2026-02-10

AI Summary

This study addresses the limited interpretability in lexical semantic change (LSC) research, which often hinders systematic understanding of semantic evolution mechanisms. The authors propose a novel approach that maps contextual word embeddings from pretrained language models into a neurobiologically grounded semantic feature space, where each dimension corresponds to an interpretable primitive semantic attribute. This mapping enables human-readable analysis of LSC by aligning computational representations with cognitively plausible semantic features. By introducing this neurobiological feature space for the first time in LSC studies, the method not only outperforms most existing approaches in estimating the degree of semantic change but also uncovers previously overlooked patterns of semantic shift. Furthermore, it facilitates targeted retrieval of words exhibiting specific types of semantic change, offering both theoretical innovation and practical utility.

Technology Category

• Natural Language Processing: Lexical Semantics and Morphology
• Machine Learning: Neuro-Symbolic Learning
• Data Mining & Knowledge Management: Semantic Web

Application Category

• 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: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Lexical Semantic Change (LSC) is the phenomenon in which the meaning of a word change over time. Most studies on LSC focus on improving the performance of estimating the degree of LSC, however, it is often difficult to interpret how the meaning of a word change. Enhancing the interpretability of LSC is a significant challenge as it could lead to novel insights in this field. To tackle this challenge, we propose a method to map the semantic space of contextualized embeddings of words obtained by a pre-trained language model to a neurobiological feature space. In the neurobiological feature space, each dimension corresponds to a primitive feature of words, and its value represents the intensity of that feature. This enables humans to interpret LSC systematically. When employed for the estimation of the degree of LSC, our method demonstrates superior performance in comparison to the majority of the previous methods. In addition, given the high interpretability of the proposed method, several analyses on LSC are carried out. The results demonstrate that our method not only discovers interesting types of LSC that have been overlooked in previous studies but also effectively searches for words with specific types of LSC.

Problem

• Lexical Semantic Change
• Interpretability
• Semantic Change
• Neurobiological Features

Innovation

• Lexical Semantic Change
• Interpretability
• Neurobiological Features
• Contextualized Embeddings
• Semantic Space Mapping