Toxic Comment Classification

Classifying a given piece of text into one of the respective toxicity categories using machine learning. Used python, scikit-learn, pandas

Team:

• Mehul Bhanushali

• Sai Krishna Pawan Suryatej Meda

• Swanand Vaishvampayan

OverView:

Digital anonymity over social media platforms have provided the liberty of expressing one’s thoughts and opinions on literally any topic. Most often these comments are inundated with unprofessional and explicit language which have adverse effects on readers. This toxicity could either suppress thoughts or invoke more hatred amongst the community. Thus, there is a dire need for a means to curb this pernicious language and provide more socially acceptable communication platforms. In this proposal we are pitching the notion of using Machine Learning algorithms to classify whether a given piece of text is toxic or nontoxic. If toxic classify them into one of the five categories Toxic, Severe_Toxic, Obscene, Threat, Insult and Identity_Hate.

Results:

Technology used:

• Python

• Sci-kit Learn

• Pandas

Learnings

• Data Preprocessing

• Advancement in Machine learnings

• Presentation