DMDD: A Large-Scale Dataset for Dataset Mentions Detection

Abstract The recognition of dataset names is a critical task for automatic information extraction in scientific literature, enabling researchers to understand and identify research opportunities. However, existing corpora mention detection are limited size naming diversity. In this paper, we introduce the Dataset Mentions Detection (DMDD), largest publicly available corpus task. DMDD consists main corpus, comprising 31,219 articles with over 449,000 mentions weakly annotated format in-text spans, an evaluation set, which comprises 450 manually purposes. We use establish baseline performance linking. By analyzing various models on DMDD, able open problems detection. invite community our as challenge develop novel models.