BabelSememe
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/thunlp/BabelNet-Sememe-Prediction
下载链接
链接失效反馈官方服务：
资源简介：
该数据集名为BabelSememe，旨在成为多语言语义知识库的基础，其中包含了针对BabelNet同义词集的手动标注语义。这些标注基于BabelNet同义词集的近义词，并且为了确保标注质量，每个目标同义词集至少由三名参与者进行标注。该数据集规模宏大，共有15,756个BabelNet同义词集被标注了43,154个语义。其研究任务是对BabelNet同义词集进行自动语义预测。
This dataset, named BabelSememe, is designed to serve as the foundation of a multilingual semantic knowledge base, and it contains manually annotated semantics for BabelNet synsets. These annotations are based on the near-synonyms of the BabelNet synsets, and to ensure annotation quality, each target synset is annotated by at least three annotators. This dataset has a substantial scale, with a total of 15,756 BabelNet synsets annotated with 43,154 semantic entries. The core research task of this dataset is to perform automatic semantic prediction for BabelNet synsets.
提供机构：
The authors of the paper
