MailFilter

Mail classification with tensorflow and MS Exchange Server

workflow

You need two mail accounts, here [email protected] (your work account) and [email protected] (collect labeled data).

1. collect train and test data

Configure quicksteps in MS Outlook ([email protected]) to change mail subjects form "bla bla bla" to "###ham### bla ..." and forward it to a new mail address ([email protected]). Do the same for spam.

2. generate csv file

Use Outlook to export mail data ([email protected]) as csv file.

3. train the ai

Use train() function to train your ai with tensorflow and store the model.

4. apply

Use apply_to_account() to classify mails ([email protected]) and move them in "ham" or "spam" subfolders in your inbox (create these folders manually in your inbox).

This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at https://mozilla.org/MPL/2.0/.