Ultimate spam detector

Presentation

Slipspam consists of a command line program that can do 3 things:

• benchmarks: bench
• predictions: predict
• email data minig: parse

bench command

bin/slipspam bench [options]

This command runs a full benchmark of every implemented algorithms over a dataset.

Here are the results of the latest benchark over the spambase dataset:

here over a spam dataset found on kaggle:

and here with spambase as the trainset and kaggle's as the testset:

predict command

bin/slipspam predict [options] (<email-text> | --in-text=<file> | --in-feat=<file>)

This command enables you to predict if a mail is spam or not. It can also run the algorithm on a full dataset via the --in-feat option.

parse command

bin/slipspam parse <in-file> <out-file>

This command convert a dataset of text mails into a dataset of features corresponding to the ones our algorithm is trained on.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

• python3
• pip3
• Tkinter

Debian

sudo apt install python3 python3-pip python3-tk

Installing

This will install and activate a virtual python environment in the dev directory.

Windows

.\activate

Unix

. activate

Running

Benchmarks

bin/slipspam bench [-e <nb executions>]

exemple:

bin/slipspam bench -e 5

Deactivating

To deactivate the virtual environment, simply type deactivate

Deployment

not deployable yet...

Documentation

CLI

You can get the CLI documentation by running the following command:

bin/slipspam --help

Built With

• scikit-learn - Machine learning library

Authors

• Anass ELANBARI - anassCoding
• Gwendal LE STANGUENNEC - GwendalLES
• Etienne LELOUËT - etienne-lelouet
• Nicolas PEUGNET - n-peugnet

See also the list of contributors who participated in this project.