Repo has flask based web app created for a usecase of classifying email as spam or not.
Welcome to SPAM FILTER application. This application can be used to perform the following tasks.
Upload a data set of labelled emails.
Train the uploaded emails dataset based on our built in Spam Classifier.
Build a Machine Learning model, capable of differentiating spam emails and non-spam emails, based on our built-in Spam Classifier.
Use the build model to predict if input emails are SPAM or NOT SPAM.
For complete usage of this application, view the rules defined below for various tasks.
Rules for Uploading a Data Set Data Set must be a csv file.
Data Set must contain two columns, with column names 'text', and 'spam' respectively.
Values in 'text' column must be email text starting with keyword 'Subject:' .
Values in 'spam' column must be either 0 or 1. 1 must indicate a SPAM email and 0 must indicate NON SPAM email.
Rules for Training and Building a Model Before Training Please upload a data set.
Choose one of the uploaded CSV files for training.
Provide inputs for parameters Test Data Set Size, Random State, and Shuffle.
A single or multiple emails can be provided as input.
Each input email must start with keyword 'Subject:' .
If multiple emails are provided, then each mail must be separated from another by atleast a single blank line.
The input can be provided either by pasting text in available area or can be submitted in a text file.
