This project aims to evaluate and compare various machine learning models for predicting consumer ratings based on their reviews of recipes from Food.com. By leveraging different approaches and techniques, the objective is to identify the most effective model that can accurately predict consumer ratings.
Food.com, formerly known as Genius Kitchen, is a popular website that allows users to share and access kitchen recipes. With 18 years of user interactions and uploads, the company has amassed a comprehensive dataset. For this project, data from Food.com was acquired via Kaggle.com, a Machine Learning and Data Science Community website. The dataset comprises 180K+ recipes and 700K+ recipe reviews.
The goal of this project is to leverage a regularized linear classifier, specifically logistic regression, to analyze the text data extracted from recipe reviews. Through cross-validation and fine-tuning of hyperparameters, such as the regularization term, the aim is to improve the baseline performance. Feature engineering and selection techniques will also be employed to enhance the model's accuracy. Finally, the project involves comparing the final model's performance with the initial baseline to assess its effectiveness in predicting consumer ratings based on recipe reviews.