SIGN
收藏arXiv2025-09-30 收录
下载链接：
http://www.philippe-fournier-viger.com/spmf/index.php?link=datasets.php
下载链接
链接失效反馈官方服务：
资源简介：
该数据集名为SIGN，是一个具有序列关系的特殊项目序列数据库，包含731个序列和267个不同的项目。每个序列的平均长度为51.997。此外，该数据集是密集型的，每个项目集合仅包含一个项目。在评估序列规则挖掘算法的性能和可扩展性任务中，算法ComSR non对该SIGN数据集的全部记录进行了测试，而由于效率问题，ComSR ful算法仅对前100条记录进行了测试。数据集的规模为731个序列和267种不同的项目。
This dataset, named SIGN, is a specialized item sequence database with sequential relationships. It contains 731 sequences and 267 distinct items, with an average length of 51.997 per sequence. Additionally, this dataset is dense, as each itemset contains only one item. For the task of evaluating the performance and scalability of sequence rule mining algorithms, the ComSR non algorithm was tested on all records of the SIGN dataset. Due to efficiency issues, the ComSR ful algorithm was only tested on the first 100 records. The scale of the dataset is 731 sequences and 267 distinct items.
