LiSCU
收藏资源简介:
LiSCU数据集由加州大学圣克鲁兹分校的Faeze Brahman创建,包含9499条文学作品摘要及其中的角色描述。数据集内容丰富,涵盖了多个文学作品的摘要和角色描述,数据来源于在线学习指南。创建过程中,使用Scrapy爬虫框架从多个在线资源中收集数据,并通过信息重叠度进行筛选和过滤。LiSCU数据集主要用于推动以角色为中心的叙事理解研究,特别是通过角色识别和角色描述生成两个任务,帮助机器更好地理解和分析文学作品中的角色及其在叙事中的作用。
The LiSCU dataset was developed by Faeze Brahman from the University of California, Santa Cruz. It contains 9,499 literary work abstracts paired with their corresponding character descriptions. The dataset covers a wide range of abstracts and character descriptions from various literary works, with its data sourced from online study guides. During the dataset creation process, the Scrapy crawler framework was employed to collect data from multiple online resources, followed by screening and filtering based on the degree of information overlap. The LiSCU dataset is primarily designed to promote character-centric narrative understanding research. Specifically, it supports two core tasks: character recognition and character description generation, aiming to assist machines in better understanding and analyzing the roles of characters in literary works and their functions within narratives.




