Toxic Comments Classifier

• 

  Toxic comment has entered and it classifies it into following categories

• 

  Here, clean comment is entered so , It shows "this comment is clean message"

Inspiration

Helps to fight Cyber Bullying

What it does

Classifies comments on Social Media into 6 categories based on there toxicity level

How I built it

Using NLP, classification of Wikipedia comments which have been labeled into 6 categories. Used spaCy for text pre-processing .Built a Logistic Regression Model for classification, got an accuracy of 96.6%. Built a Web-app using Streamlit and deployed it on Heroku.

Challenges I ran into

It was really difficult to host it into Heroku as the libraries which I used was not supported by Heroku, So I have to add some extra files.

Accomplishments that I'm proud of

I am proud of that my Logistic regression Model got an accuracy of 97%

What I learned

I learned Natural Language Processing, how o pre-process texts and tokenise it using spaCy.

What's next for Toxic Comments Classifier

I am looking forward for continuous improvements by using word2vec technology.

Built With

• heroku
• matplotlib
• numpy
• pandas
• python
• sklearn
• spacy
• streamlit
• tensorflow