COS 781 Project - Richard Lastrucci

This repository contains the code and data for my COS 781 Project, focusing on building a recommendation system. Below is an overview of the project structure, data flow, and usage instructions.

---

📂 Project Structure

1. Data

• Located in the data folder.
• Excludes the raw data and intermediate data due to size, please contact me if you wish to have them
• Includes the preprocessed and application datasets.
• The workflow starts with the source data and progresses through intermediate, preprocessed, and final application stages.

2. Code

The codebase is organized into the following key folders:

Preprocessing

Contains three Jupyter notebooks:

• source.ipynb: Handles data loading and sampling.
• preprocess.ipynb: Includes data subsetting, cleaning, exploratory data analysis (EDA), and utility matrix construction.
• clustering.ipynb: Performs clustering analysis on the normalized utility matrix.

Application

Contains two Jupyter notebooks for applying recommendation algorithms:

• cosine_similarity.ipynb: Implements cosine similarity on the utility matrix.
• matrix_factorization.ipynb: Implements matrix factorization on the utility matrix.

Helpers

• Contains styling functions and plot utilities to enhance visualizations and readability.

---

Environment

The env.yml file specifies the Conda environment required to run the code. To set up the environment: