Comment Toxicity Classification

This model helps to predict the is a comment/sentence is hateful on various parameters such as toxicity, severe toxicity, obscene, threat, insult and racism.

Test the model here : pvcodes/comment_toxicity_classifier

Working of the Model

• Loading of Data

  The data is fetched from csv file, which consist of the comment and attributes such as toxicity, severe toxicity, obscene, threat, insult and racism.

• Preprocessing the comments

  Then the data is tokenized using the TextVectorization method of keras in and embeded

• Creating of Deep NLP Model

  For this model we used Keras sequential API with a number of LSTM layers (because they are particulary good while working with sequences)
• The dataset used to train the model is from <a href=https://www.kaggle.com/c/jigsaw-toxic-comment-classification-challenge>Toxic Comment Classification Challenge from <a href=https://www.kaggle.com>Kaggle.

Note: The compiled data model is available here: here.

════ ⋆★⋆ ════
From pvcodes