Scientific Literature
Scientific literature analysis is undergoing a transformation driven by advancements in natural language processing (NLP) and large language models (LLMs). Current research focuses on automating tasks like information extraction, summarization, and citation generation using various architectures, including graph neural networks and transformer-based models, to improve accessibility and efficiency in navigating the vast and growing body of scientific publications. This work aims to enhance knowledge discovery, accelerate research, and improve the reliability and transparency of scientific findings, with applications ranging from drug discovery to policy-making. The development of robust benchmarks and datasets is crucial for evaluating and improving these methods.
Papers
Automating Knowledge Discovery from Scientific Literature via LLMs: A Dual-Agent Approach with Progressive Ontology Prompting
Yuting Hu, Dancheng Liu, Qingyun Wang, Charles Yu, Heng Ji, Jinjun Xiong
Towards Efficient Large Language Models for Scientific Text: A Review
Huy Quoc To, Ming Liu, Guangyan Huang