/wine-sentiment

A statistical algorithm for sentiment-based classification of wine reviews based on word counts.

Primary LanguageJupyter Notebook

Wine Review Sentiment Analysis

I created a statistical algorithm for sentiment-based classification of wine reviews based on word counts.

Features

  • an algorithm created from scratch using only Python, Pandas and MatPlotLib
  • positive/negative scoring of terms based on their relative frequencies in both sentiment categories
  • filtering a subset of terms to only count the most important ones in each review
  • achieved ~95% accuracy on test set

Requirements

This project only requires Python, Pandas and MatPlotLib. You will also need to download the dataset and extract it to the same directory as the Jupyter notebook.