Enhancing Unsupervised Sentence Embedding through Knowledge-Based Domain-Oriented Data Augmentation
A novel pipeline-based data augmentation method that leverages large language models to synthesize domain-specific datasets, enhancing fine-grained sentence representation learning.