Diachronic semantic encoding based on pre-trained language model for temporal knowledge graph reasoning

Yunteng Deng, Jia Song, Zhongliang Yang, Yilin Long, Li Zeng, Linna Zhou

Abstract: Highlights•Address the problem of lack of semantic awareness in traditional GNN-based models.•Introduce pre-trained language models to enhance the understanding of historical facts.•Use the semantic encoding information of entities and relations as the initial input.