Named Entity Recognition (NER) is a fundamental task in processing biomedical text. Due to the limited availability of labeled data, researchers have investigated few-shot learning methods to tackle this challenge. However, replicating the performance of fully supervised methods remains difficult in few-shot scenarios. This paper addresses two main issues. In terms of data augmentation, existing methods primarily focus on replacing content in the original text, which can potentially distort the semantics. Furthermore, current approaches often neglect sentence features at multiple scales. To overcome these challenges, we utilize ChatGPT to generate enriched data with distinct semantics for the same entities, thereby reducing noisy data. Simultaneously, we employ dynamic convolution to capture multi-scale semantic information in sentences and enhance feature representation based on PubMedBERT. We evaluated the experiments on four biomedical NER datasets (BC5CDR-Disease, NCBI, BioNLP11EPI, BioNLP13GE), and the results exceeded the current state-of-the-art models in most few-shot scenarios, including mainstream large language models like ChatGPT. The results confirm the effectiveness of the proposed method in data augmentation and model generalization.

The published biomedical literature contains extensive amount of valuable undiscovered knowledge and hidden relations. Classical information retrieval techniques allow to access explicit information from a given collection of information, but are not able to recognize implicit connections. Literature-based discovery (LBD) is characterized by uncovering hidden associations in non-interacting literature. It could significantly support scientific research by identifying new connections between biomedical entities. However, most of the existing approaches are not applicable to both closed and open discovery tasks. Here, we present a model which incorporates knowledge graph, graph embedding and deep learning methods for both open and closed LBD. In this work, we showed how deep learning combining with graph embedding techniques can be applied to LBD tasks, including discovering new knowledge from unrelated literature (open LBD), and providing logical explanations for the relations between entities (closed LBD). The experimental results suggest that incorporating knowledge graph and deep learning methods is an effective way for capturing the underlying complex associations between entities hidden in the literature.