Gourmet: An Automated approach to recommendation engines - Project Team 14
Abstract— the aim of this paper is to create a scalable API bases recommendation engine. The project is built using technologies such as Node.js, Python, Amazon AWS, Machine learning etc. The project is deployed on Heroku for 24X7 availability. The project reads a huge data set of 70,000 recipes of various cousins and using the 80-20 rule it trains itself with the 80% of the data set (Training Set) and builds a multidimensional mathematical clustering model of the data set using python libraries. The model is then trained with the training set. After sufficient training the model is tested with the test set for generating the recommendations for various recipes based of the ingredients.
As food is the basic need and with changing times it is taking the form of art so in this project we want to use machine learning to approach food and specifically recipes in new ways. For this we will classify the cuisine in categories and taking input from user what all ingredients user have at the moment and our algorithm will give options to the users what all user can cook and what cuisine it belongs to if not selected already. Also, we will include features like what all are the nutrient contents of the selected recipe and calories intake etc. and recommend what all user can also cook if they have missing ingredients, basically in order to give more options if user can arrange few additional ingredients.
Keywords— Recommendation Engine, Machine Learning, API, AWs, Clustering,
Link to our Project: http://34.208.105.157:8080/homepage
PPT: https://github.com/SJSU272LabS17/Project-Team-14/blob/master/Team14-cmpe272.pdf
Members:
- Deepika Kalani
- Sunil Tiwari
- Kanika Gupta
- Swati Gupta