Goodreads Book Reading Histories

该数据集是从Goodreads提取的书籍阅读历史记录，涵盖了超过26,000名用户和近130万本书。其中，针对4名用户的449本书籍阅读记录进行了手动标注，标注内容涉及这些书籍随时间变化的话题级惊喜感。此外，该数据集还包括基于话题级分析，针对4名参考用户样本的书籍惊喜感的手动标注信息。规模上，数据集覆盖了超过26,000名用户和大约130万本书籍。该数据集的任务是评估推荐系统中的话题级模型在惊喜和偶然发现方面的表现。

This dataset is extracted from Goodreads book reading history records, covering over 26,000 users and nearly 1.3 million books. Among them, 449 book reading records of 4 users have been manually annotated, with the annotations involving the topic-level surprise of these books over time. Additionally, this dataset also includes manually annotated information about book surprise levels for 4 reference user samples based on topic-level analysis. In terms of scale, this dataset covers more than 26,000 users and approximately 1.3 million books. The task of this dataset is to evaluate the performance of topic-level models in recommendation systems with regard to surprise and serendipitous discovery.