用于匿名化的PYTHON库

此库支持表格数据应用三种经典匿名化技术：k-匿名、l-多样性和t-接近度。

安装

我们建议使用Python3和virtualenv

> virtualenv .venv -p python3
> source .venv/bin/activate

然后运行以下命令安装库及其所有依赖项

pip install python-anonymity

文档

python-anonymity文档托管在Read the Docs上。

入门指南

使用犯罪合成数据集的示例

> import pandas as pd
> import pycanon
> from anonymity import tools
> from anonymity.tools.utils_k_anon import utils_k_anonymity as utils
> 
> d = {
>         "name": ["Joe", "Jill", "Sue", "Abe", "Bob", "Amy"],
>         "marital stat": [
>             "Separated",
>             "Single",
>             "Widowed",
>             "Separated",
>             "Widowed",
>             "Single",
>         ],
>         "age": [29, 20, 24, 28, 25, 23],
>         "ZIP code": ["32042", "32021", "32024", "32046", "32045", "32027"],
>         "crime": ["Murder", "Theft", "Traffic", "Assault", "Piracy", "Indecency"],
>     }
>     data = pd.DataFrame(data=d)
> 
>     ID = ["name"]
>     QI = ["marital stat", "age", "ZIP code"]
>     SA = ["crime"]
>     age_hierarchy = {"age": [0, 2, 5, 10]}
>     hierarchy = {
>         "marital stat": [
>             ["Single", "Not married", "*"],
>             ["Separated", "Not married", "*"],
>             ["Divorce", "Not married", "*"],
>             ["Widowed", "Not married", "*"],
>             ["Married", "Married", "*"],
>             ["Re-married", "Married", "*"],
>         ],
>         "ZIP code": [
>             ["32042", "3204*", "*"],
>             ["32021", "3202*", "*"],
>             ["32024", "3202*", "*"],
>             ["32046", "3204*", "*"],
>             ["32045", "3204*", "*"],
>             ["32027", "3202*", "*"],
>         ],
>     }
> 
>     mix_hierarchy = dict(hierarchy, **utils.create_ranges(data, age_hierarchy))

>     k = 2
>     supp_threshold = 0
>     new_data = tools.data_fly(data, ID, QI, k, supp_threshold, self.mix_hierarchy)
> 

许可证： Apache 2.0.

注意：此库正在积极开发中，仅供测试目的。