Command-line Logs

该数据集包含了从大约10万台机器的用户处收集的命令行日志，数据覆盖了一周的时间。其中，训练集样本量达到了3000万个，测试集样本量为1000万个。为了便于评估指标，数据集中包含了去重后的样本。原始的训练文件和测试文件分别占用了10.5GB和4.6GB的磁盘空间。该数据集的任务是入侵检测。

This dataset contains command-line logs collected from users across approximately 100,000 machines, covering a one-week period. The training set comprises 30 million samples, while the test set contains 10 million samples. To facilitate metric evaluation, the dataset includes deduplicated samples. The original training and test files occupy 10.5 GB and 4.6 GB of disk space, respectively. The targeted task of this dataset is intrusion detection.