EICAT dataset
收藏资源简介:
EICAT数据集是一个针对入侵物种对生态系统影响评估的新数据集,包含特定入侵物种的全文科学论文和相应的证据句子,这些论文与IUCN的EICAT标准相一致。数据集由比勒费尔德大学计算语言学系创建,包含436篇全文,针对120个物种,通过模糊匹配策略将证据句子与全文文本匹配。该数据集可用于训练和评估模型在基于科学全文的入侵物种影响评估任务。
The EICAT dataset is a novel dataset for assessing the impacts of invasive species on ecosystems. It contains full-length scientific papers on specific invasive species and their corresponding evidence sentences, which conform to the IUCN EICAT standards. Developed by the Department of Computational Linguistics at Bielefeld University, the dataset includes 436 full texts covering 120 species, with evidence sentences matched to the full texts through fuzzy matching strategies. This dataset can be utilized to train and evaluate models for the task of invasive species impact assessment based on full scientific papers.




