Let’s assume a collection of academic papers and a group of researchers. The citation recommendation model assesses the potential advantage a distinct paper from this collection may offer to a specific researcher. This assessment is conducted through a utility function, designated to quantify the benefit. Essentially, this function ascertains the degree of alignment between a paper and a researcher’s requirements or inclinations.
A recent paper by Yang Zhang et al. (DOI: 10.48550/arXiv.2309.09727) summarises existing research on how Large Language Models (LLMs) contribute to citation-related tasks. The picture in this post represents their conceptual framework. There are exciting opportunities in this space for new developments.
