Manage your emails with Python

The pydatamail_ml python module extends the functionality of pydatamail by adding support for machine learning. This includes but is not limited to text preprocessing utilities as well as basic machine learning models based on decision trees.

Installation

Install the pydatamail_ml package using pip:

pip install pydatamail_ml

Or alternatively using conda:

conda install -c conda-forge pydatamail_ml

Python interface

Import the pydatamail_ml module

from pydatamail_ml import one_hot_encoding, get_machine_learning_database, gather_data_for_machine_learning, train_model, get_machine_learning_recommendations, text_pipeline, detect_language

The individual components are briefly explained below:

• one_hot_encoding() - Encoding the email headers fields (to, from, cc) in binary descriptors.
• get_machine_learning_database() - Get a database to store machine learning models in.
• get_gather_data_for_machine_learning() - Clean up encoded data by separating input and output.
• train_model() - Train a randomforest machine learning model.
• get_machine_learning_recommendations() - Get recommendations from the machine learning model.
• text_pipeline() - Convert text from the email to machine readable text by removing HTML design elements.
• detect_language() - Detect the language of the email.