Diachronic Semantic Evolution Detection and Attribution Analysis Based on Large Language Models: Taking Chinese Emotional Words as an Example
Abstract
The diachronic semantic evolution of Chinese emotional words is an important mirror image of language and social and cultural changes. Traditional methods have limitations such as high corpus dependence and insufficient mechanism interpretation. This research proposes a detection and attribution framework based on large language models, constructs a diachronic corpus covering multiple sources such as newspapers, periodicals, and online texts from 1995 to 2025, and uses large language models to quantify the semantic evolution of emotional words precisely. And by incorporating external attributions such as social events, media discourse, and cultural trends, the research reveals how multiple factors jointly drive semantic evolution. The results show that over the past 30 years, the evolution of Chinese emotional words has presented significant stages, and their core meanings and practical contexts have changed dynamically. The research breaks through the bottleneck of traditional methodology, providing an innovative computational paradigm for diachronic semantic research, enriching the theory of language and social interaction, and also offering practical technical support for the construction of Chinese diachronic language resources and the exploration of historical culture.
References
- Shen, J. X. (2009). Introduction to Cognitive Linguistics. The Commercial Press.
- Wang, Q. (2025). AETAS: Analysis of Evolving Temporal Affect and Semantics for Legal History. arXiv preprint arXiv: 2512.22196.
- Chen, J., Chersoni, E., Schlechtweg, D., Prokić, J., & Huang, C. R. (2023, December). ChiWUG: A graph-based evaluation dataset for Chinese lexical semantic change detection. In Proceedings of the 4th Workshop on Computational Approaches to Historical Language Change (pp. 93-99).
- Dhingra, B., Cole, J. R., Eisenschlos, J. M., Gillick, D., Eisenstein, J., & Cohen, W. W. (2022). Time-aware language models as temporal knowledge bases. Transactions of the Association for Computational Linguistics, 10, 257-273.
- Giulianelli, M., Del Tredici, M., & Fernández, R. (2020). Analysing lexical semantic change with contextualised word representations. arXiv preprint arXiv: 2004.14118.
- Qiu, W., & Xu, Y. (2022). Histbert: A pre-trained language model for diachronic lexical semantic analysis. arXiv preprint arXiv: 2202.03612.
- Hamilton, W. L., Leskovec, J., & Jurafsky, D. (2016). Diachronic word embeddings reveal statistical laws of semantic change. arXiv preprint arXiv: 1605.09096.
- Ren, Z., Caputo, A., & Jones, G. (2024, August). A few-shot learning approach for lexical semantic change detection using GPT-4. In Proceedings of the 5th Workshop on Computational Approaches to Historical Language Change (pp. 187-192).
- Kang, C., Baek, J., Kim, Y., & Choi, J. (2025, November). When Format Changes Meaning: Investigating Semantic Inconsistency of Large Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2025 (pp. 2647-2667).
- Tahmasebi, N., Borin, L., & Jatowt, A. (2021). Survey of computational approaches to lexical semantic change detection. Computational approaches to semantic change, 6(1).
- Karjus, A. (2023). Machine-assisted quantitizing designs: augmenting humanities and social sciences with artificial intelligence. arXiv preprint arXiv: 2309.14379.
- Azarbonyad, H., Dehghani, M., Beelen, K., Arkut, A., Marx, M., & Kamps, J. (2017, November). Words are malleable: Computing semantic shifts in political and media discourse. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management (pp. 1509-1518).