Recommendation System Using Factorization Model and MapReduce Framework

This project is part of my master's thesis 👨‍🎓 where we developed a recommendation system using factorization model and MapReduce framework to predict the rating of an item by a user. The factorization model is trained using stochastic gradient descent and is deployed in Amazon EMR (Amazon Elastic MapReduce) cluster for distributed computing.

The code is organized into two directories: train and predict. Each of the directory contains README file that describes the files present in them.

• train : contains program to train the model, further details are in README file inside the train directory.
• predict : contains program to predict the rating using MRJob, further details are in README file inside the predict directory.

Installation and Usage

1. Clone the repository to your local machine
2. Install the required packages
3. If you wish to train the model, please follow the instructions in the README file inside the train directory
4. If you wish to use the trained model for prediction, please follow the instructions in the README file inside the predict directory

Availability

To dive deeper into my master's research thesis and explore the methods, please read the full thesis available online. And for a quick and engaging overview of my research, we invite you to watch my three-minute thesis presentation.

• To read the thesis, please visit -> Zenodo Record.
• To watch the three-minute thesis presentation, please visit -> 3MT Video in YouTube.

Contributors

This project was developed as a part of a master's thesis by Nabin Giri under the guidance of Dr. Yui Man Lui at the University of Central Missouri, Warrensburg, Missouri.

Citation

We are passionate about our research and committed to driving innovation in our field. If you find our project or any of its elements useful in your own work or research, we kindly ask that you consider citing it appropriately. Your support and recognition of our efforts help to drive our research forward and make a positive impact in the field.

We welcome discussions and collaborations, so please do not hesitate to reach out if you have any questions or ideas to share. Thank you for your interest in our work!

@mscthesis{giri_nabin_2020_6591586,
  author       = {Giri, Nabin},
  title        = {{Recommendation System Using Factorization Model 
                   and MapReduce Framework}},
  school       = {University of Central Missouri},
  year         = 2020,
  month        = May,
  note         = {{Three-minute thesis presentation available at : 
                   https://youtu.be/KVL9eQ35YSY}},
  doi          = {10.5281/zenodo.6591586},
  url          = {https://doi.org/10.5281/zenodo.6591586}}