Meal Magic

Inspiration

Often times it can be very difficult and time consuming to choose where to go out to eat with family, friends, or coworkers. Team 3 set out to simplify this process and generate excellent restaurant recommendations by using the nearest neighbor algorithm.

What it does

After taking a short survey and evaluating 40 Philadelphia restaurants on a scale from 1 and 5, user input their name and the number of recommendations and similar users they would like to receive. Meal Magic looks the user up in the system and then outputs a list of restaurant recommendations and similar users personalized to the preferences of the current user.

How we built it

Both the algorithm and user interface for Meal Magic are written and executed in Matlab.

Challenges from first semester

Finding the perfect framework and language in which to write Meal Magic's nearest neighbor algorithm proved to be a very difficult task. Team 3 pivoted between Python, Matlab, and Apache Mahout over a 3-month period, and ended up deciding to move forward with Matlab to build the product for the Fall Demo Day.

What's next for Meal Magic

Next semester, Team 3 will turn Meal Magic into a dynamic web-app written in Python. Existing users will be able to dynamically and remotely edit their preferences, and new users will be able to create accounts to receive their own recommendations.

Built With

• matlab