A survey on knowledge graph evolution: proliferation, dynamic embedding, and versioning

In the era of large language models (LLMs), knowledge graphs (KGs) can play a pivotal role in enhancing LLMs by providing a structured representation of knowledge, relationships, and entities. This knowledge is essential for LLMs to understand and interpret information in a coherent and contextually relevant manner. KGs must undergo continuous evolution with minimal human intervention to remain effective. Organisations often employ automated techniques, such as web scraping, natural language processing, and machine learning algorithms, to accomplish this continuous evolution. However, there is a lack of reviews covering recent advances in KG evolution. In this survey, we first give an overview and then describe the methods of KG evolution. Afterward, we review and analyse the evaluation metrics, datasets, and experimental performances. Finally, we provide findings and future directions from the investigation and conclude with a discussion.