Skip to content

ekapopev/IMDb_Sentiment_Analysis

Repository files navigation

IMDB Movie Reviews Sentiment Analysis

Authors:

Viriyakovithya Ekapope | Chauhan Siddhant | Ponce James | Piracha Ali

Project Summary:

We created a shiny app to do sentiment classification (Positive / Negative) on IMDB movie review dataset. Classification is done through a neural network model (trained with 8 special features engineered from movie reviews) that performs with an AUC of 91% (10 pts improvement vs. base model).

For more detail, please refer to the report.

Apart from the classification task the shiny app also provides visualizations that generate insights into the test reviews data.

You can try by uploading your own dataset directly in ShinyApp. Sample csv input can be found here. The ShinyApp is published via this link.