/flask-spam-classifier

Primary LanguageHTML

Spam Classifier

A project for classifying whether the email/text received is Spam or Ham. I have used here Naive Bays classifier for training the model and deployed using Flask. You can have a look of the working project from the link below.

Published Site

https://flask-email-classifier.herokuapp.com/

Build With

  • Python
  • ML (used Naive Bays classifier)
  • Flask
  • HTML, CSS, JS

In this project I learnt and worked on important steps like: 
  πŸ”Έ 𝑬𝑫𝑨 (π‘¬π’™π’‘π’π’π’“π’‚π’•π’π’“π’š 𝒅𝒂𝒕𝒂 π’‚π’π’‚π’π’šπ’”π’Šπ’”)    
          
  πŸ”Έ 𝑫𝒂𝒕𝒂 π‘·π’“π’†π’‘π’“π’π’„π’†π’”π’”π’Šπ’π’ˆ
          - Text Cleaning Procedures
              ⚬  converting all words in document to lower case
              ⚬  Tokenizing
              ⚬  Removing stop words
              ⚬  Word stemming
              ⚬  Word lemmatization
              ⚬  Removing punctuations
              ⚬  Stripping out HTML tags
          ππ‹π“πŠ library was there to think out of the box!
              
  πŸ”Έ 𝑭𝒆𝒂𝒕𝒖𝒓𝒆 π‘¬π’™π’•π’“π’‚π’„π’•π’Šπ’π’
          - CountVectorizer Method
          - Full Matrix Creation
          - Vocabulary Creation
          
  πŸ”Έ π‘¨π’π’ˆπ’π’“π’Šπ’•π’‰π’Ž π‘°π’Žπ’‘π’π’†π’Žπ’†π’π’•π’‚π’•π’Šπ’π’
          - CountVectorizer + NaΓ―ve Bayes Algorithm
          
  πŸ”Έ π‘Ίπ’„π’π’“π’Šπ’π’ˆ & π‘΄π’†π’•π’“π’Šπ’„π’” :
          - Accuracy
          - Precision
          - Recall

image image