MOTIF

MOTIF数据集是由马里兰大学巴尔的摩分校和Booz Allen Hamilton合作创建的，包含3095个恶意软件样本，来自454个不同的家族，是目前最大的公开恶意软件数据集，具有专家验证的家庭标签。数据集中的每个样本都与来自可信行业来源的威胁报告相关联，这些报告不仅验证了标签，还为研究提供了新的机会。MOTIF数据集特别适用于评估恶意软件家族分类工具的准确性，尤其是在处理来自不同来源的名称时。此外，数据集还包含了恶意软件家族的全面别名映射，这使得能够进行以前无法执行的重要评估。MOTIF数据集的应用领域包括但不限于恶意软件家族分类、恶意软件行为分析和恶意软件检测技术的评估。

The MOTIF dataset, co-created by the University of Maryland, Baltimore and Booz Allen Hamilton, contains 3,095 malware samples spanning 454 distinct families. It is currently the largest publicly available malware dataset with expert-validated family labels. Each sample in the dataset is linked to threat reports from trusted industry sources, which not only validate the labels but also provide new research opportunities. The MOTIF dataset is particularly well-suited for evaluating the accuracy of malware family classification tools, especially when dealing with names from disparate sources. Additionally, the dataset includes comprehensive alias mappings for malware families, enabling critical evaluations that were previously infeasible. Application areas of the MOTIF dataset include, but are not limited to, malware family classification, malware behavior analysis, and the evaluation of malware detection technologies.

搜集汇总

背景与挑战

背景概述

MOTIF数据集是一个大型恶意软件参考数据集，包含3,095个无害化PE格式恶意软件样本，覆盖454个家族，并提供家族标签、别名映射和EMBER原始特征。数据集还包含详细的报告信息、恶意软件样本的哈希值以及用于机器学习的特征向量和标签，适用于恶意软件家族分类和威胁情报研究。