Skip to content

The Enron-Spam dataset preprocessed in a single, clean csv file.

License

Notifications You must be signed in to change notification settings

MWiechmann/enron_spam_data

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Enron Spam Dataset

The Enron-Spam dataset is a fantastic ressource collected by V. Metsis, I. Androutsopoulos and G. Paliouras and described in their publication "Spam Filtering with Naive Bayes - Which Naive Bayes?". The dataset contains a total of 17.171 spam and 16.545 non-spam ("ham") e-mail messages (33.716 e-mails total). The original dataset and documentation can be found here.

However, the original datasets is recorded in such a way, that every single mail is in a seperate txt-file, distributed over several directories. This can make reading in the data a bit cumbersome, especially for beginners. Since the data set is such an excellent ressource, I wanted to create a offer a single download of the data through a simple csv-file.

You probably only need the data file ("enron_spam_data.zip"). The python file ("build_data_file.py") contains the script used to construct the csv file (downloading the original raw data from the website, unpacking it, processing it and saving it into a compressed csv file).

Processing of the data is minimal: The dataset contains the following columns:

Column Explanation
Subject The subject line of the e-mail
Message The content of the e-mail. Can contain an empty string if the message had only a subject line and no body. In case of forwarded emails or replies, this also contains the original message with subject line, "from:", "to:", etc.
Spam/Ham Has the values "spam" or "ham". Whether the message was categorized as a spam message or not.
Date The date the e-mail arrived. Has a YYYY-MM-DD format.

About

The Enron-Spam dataset preprocessed in a single, clean csv file.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages