ML Classification Pipeline for DomainRadar

Requirements

• Python 3.11
• poetry 1.8 (Please install using pipx)

Note: If you have a newer default Python version, make sure this venv uses 3.11, e.g.: poetry env use /usr/local/bin/python3.11

Usage

Modify the pyproject.toml file in your Python Poetry project - extend the [tool.poetry.dependencies] section with: classifiers = { git = git@github.com:nesfit/domainradar-clf.git, branch = "main }

Then type: poetry update

In your code, you can use the module like:

import pandas as pd
from classifiers import pipeline

# Initialize the classification pipeline
p = Pipeline()

# Load or prepare a dataframe with feature vectors
df = pd.DataFrame({
    "domain_name": ['first.com', 'second.net', ...],
    "lex_name_len": [9, 10, ...],
    ...
})

# Classify!
results = p.classify_domains(df)
print(results)

See example.py for a working example.

NOTE: Don't forget to run you code from poetry shell.

Development

Install Python Poetry. In the domainradar-clf root directory, enter:

poetry install

To experiment with the included notebooks, you have to explicitly include the development dependencies:

poetry install --with dev

You can run scripts from the classifiers directory from

poetry shell