The Creation of ArSyTa, a Novel 8.27-Million-Context Dataset for Local Citation Recommendation

This paper details ArSyTa, a new, massive dataset for citation AI. It has 8.27M rich contexts from arXiv papers to improve recommendation models.