cyberbullying-tweet-prediction-app

Web app Link: https://apurvayadav-cyberbullying-tweet-recognition-app-webapp-z6y643.streamlitapp.com/

This app predicts the nature of the tweet into 6 Categories.

• Age
• Ethnicity
• Gender
• Religion
• Other Cyberbullying
• Not Cyberbullying

Tools and Technologies

Model: Used linear Support Vector Machine to classify tweets.
Streamlit: Used Streamlit to create simple webapp to make the model interactive.

Methodology

• Downloaded the data from kaggle. (data)
• Performed some Exploratory Data Analysis to get the overview of data. (initial_modelling.ipynb)
• Created a Word Cloud from the data.
• Performed the necessary steps for textual analysis.
  • Removing Stopwords, puctuations, URLs, etc
  • Performed Stemming and Lemmatization.
• Automated the process of preprocessing by creating functions. Which would be helpful in predicting Custom Outputs.
• Created Illustrations for the webapp from Canva. (Images)
• Deployed the webapp on streamlit.

References

• Twitter Sentiment Analysis- A NLP Use-Case for Beginners - https://www.analyticsvidhya.com/blog/2021/06/twitter-sentiment-analysis-a-nlp-use-case-for-beginners/