Continuous Learning for Small-Sample Entity Extraction Based on Self-Training and Data Augmentation

Entity extraction aims to automatically identify and extract entities with specific meanings from text. It is a fundamental and crucial step in the construction of knowledge graphs, and its performance directly impacts the overall quality of the knowledge graph. Given its significant role in knowledge graph construction, this paper addresses two key issues with traditional entity extraction models: reliance on large amounts of annotated data and inability to continuously learn new emerging entity types. To tackle these problems, we propose a method for small-sample entity extraction under continuous learning based on self-training and data augmentation. This method effectively enhances the performance of small-sample entity extraction models in a continuous learning context.