Related Work Generation

Automatic related work generation (RWG) aims to automate the creation of literature review sections in academic papers, a crucial task for establishing research novelty. Current research focuses on leveraging large language models and advanced techniques like contrastive learning and causal intervention to generate coherent and informative summaries from full-text papers, moving beyond previous limitations of using only abstracts. This automated process has the potential to significantly accelerate the research workflow, improving efficiency and allowing researchers to focus on higher-level tasks. Challenges remain in accurately evaluating the quality of generated text and ensuring the generated content accurately reflects the nuances of the cited works.