To view the full report, please visit here.

Personalization is getting more and more crucial for user experiences on various websites. In this paper, we first performed exploratory data analysis on datasets from food.com. We then implemented various types of recommendation system models to recommend recipes to users, predict ratings based on sentiment analysis, and predict recipe categories. We have a baseline and an improved model for each predictive task for comparison purposes. After fine-tuning the hyperparameters for our improved models, we achieved 73.87% testing accuracy on the recipe recommendation model, 0.96 testing MSE on the rating predictions, and 94.43% testing accuracy on the category predictions. We further propose future research directions to explore and validate the causes of such differences between the baseline models and the improved models.

We have provided a requirements.txt.

All source code could be found in Recipe-Recommender.ipynb.