Amazon Access示例数据集

Dataset creator and donator: Ken Montanez email: kenmonta[at]cal.berkeley.edu institution: Information Security, Amazon Corp. Data Set Information: This is a sparse data set, less than 10% of the attributes are used for each sample. The link is to a '*.tgz' file which contains two files: [amzn-anon-access-samples-2.0.csv] this file contains the access for users [amzn-anon-access-samples-history-2.0.csv] this file contains the access history for a given user Attribute Information: __amzn-anon-access-samples-2.0.csv__ This is a sparse data set containing users and their assigned access. The file contains 4 categories of attributes. 1) [PERSON_{ATTRIBUTE}] This category describes the 'user' who was given access. The [PERSON_ID] column is the primary key column for the file. There is one row per user. PERSON_ID: id of the user PERSON_MGR_ID: id of the user's manager PERSON_ROLLUP_1: user grouping id PERSON_ROLLUP_2: user grouping id PERSON_ROLLUP_3: user grouping id PERSON_DEPTNAME: department desciption id PERSON_LOCATION: region id PERSON_BUSINESS_TITLE: title id PERSON_BUSINESS_TITLE_DETAIL: description id PERSON_JOB_CODE: job code id PERSON_COMPANY: company id PERSON_JOB_FAMILY: job family id 2) [RESOURCE_{ID}] This category of attributes are the resources that a users can possibly have access to. A user will have a 1 in this column if the have access to it otherwise it will be 0. 3) [GROUP_{ID}] - This category of attributes are the groups that a users can possibly have access to. A user will have a 1 in this column if the have access to it otherwise it will be 0. 4) [SYSTEM_SUPPORT_{ID}] - This category of attributes are the system that a user can possibly be supporting. A user will have a 1 in this column if the have can possibly be supporting it, otherwise it will be 0. __amzn-anon-access-samples-history-2.0.csv__ Permissions Time series data. Here is a short description of the columns: ACTION: either 'remove_access' or 'add_access' TARGET_NAME: either the {RESOURCE_ID} or {GROUP_ID} LOGIN: the id of the user that is obtaining or losing access REQUEST_DATE: YYYY-MM-DD HH:MM:SS AUTHORIZATION_DATE: YYYY-MM-DD HH:MM:SS Relevant Papers: N/A Citation Request: Please refer to the Machine Learning Repository's citation policy.

数据集创建者与捐赠者：Ken Montanez，邮箱：kenmonta[at]cal.berkeley.edu，所属机构：亚马逊公司（Amazon Corp.）信息安全部门。

数据集概况：本数据集为稀疏数据集，每个样本仅使用不足10%的属性。提供的链接指向一个*.tgz压缩包，内含两个文件：
1. [amzn-anon-access-samples-2.0.csv]：存储用户及其被分配的权限数据
2. [amzn-anon-access-samples-history-2.0.csv]：存储指定用户的权限历史数据

属性说明：
### __amzn-anon-access-samples-2.0.csv__
该稀疏数据集包含用户及其被分配的权限信息，共包含四类属性：
1. **[PERSON_{ATTRIBUTE}] 用户相关属性**：用于描述被授予权限的用户主体。其中[PERSON_ID]列为该文件的主键（primary key），每个用户对应一行数据。各字段详情如下：
- PERSON_ID：用户ID
- PERSON_MGR_ID：用户直属经理ID
- PERSON_ROLLUP_1：用户分组ID
- PERSON_ROLLUP_2：用户分组ID
- PERSON_ROLLUP_3：用户分组ID
- PERSON_DEPTNAME：部门描述ID
- PERSON_LOCATION：区域ID
- PERSON_BUSINESS_TITLE：职位名称ID
- PERSON_BUSINESS_TITLE_DETAIL：职位详情ID
- PERSON_JOB_CODE：岗位代码ID
- PERSON_COMPANY：公司ID
- PERSON_JOB_FAMILY：岗位族ID
2. **[RESOURCE_{ID}] 资源相关属性**：表示用户可访问的资源。若用户拥有该资源的访问权限，则对应列取值为1，否则为0。
3. **[GROUP_{ID}] 用户组相关属性**：表示用户可访问的用户组。若用户拥有该用户组的访问权限，则对应列取值为1，否则为0。
4. **[SYSTEM_SUPPORT_{ID}] 系统支持相关属性**：表示用户可能负责支持的系统。若用户可支持该系统，则对应列取值为1，否则为0。

### __amzn-anon-access-samples-history-2.0.csv__
该文件存储权限时序数据，各字段简要说明如下：
- ACTION：操作类型，仅包含'remove_access'（移除权限）与'add_access'（添加权限）两种取值
- TARGET_NAME：目标对象，可为{RESOURCE_ID}或{GROUP_ID}
- LOGIN：获取或丧失权限的用户ID
- REQUEST_DATE：权限请求时间，格式为YYYY-MM-DD HH:MM:SS
- AUTHORIZATION_DATE：权限授权时间，格式为YYYY-MM-DD HH:MM:SS

相关论文：无（N/A）
引用要求：请遵循机器学习仓库（Machine Learning Repository）的引用规范。

搜集汇总

背景与挑战

背景概述

Amazon Access示例数据集是一个稀疏数据集，包含用户及其分配的访问权限信息，主要分为用户属性、资源访问、组访问和系统支持四个类别。数据集由两个文件组成，分别记录用户的访问权限和访问历史，适用于业务回归分析场景。