The food schedules of Rice's serveries are not as easy to access as they should be. Students usually don't know what food which servery has, and thus are not able to go to the one best suited to them. Moreover, students with allergies or other dietary requirements sometimes have trouble finding the servery that would accommodate their needs.

What it does

Servery Hacks fetches each servery's food information from every day and use our algorithm to recommend you the serveries that would best fulfill your dietary restrictions and preferences.

How we built it

We first designed the core part of this project: the recommendation system. We evaluated different methods of matching and recommending menus to users with different preferences and settled with combining Google Custom Search and Wikipedia to obtain relevant information about the menus. We then started working on the code for that algorithm, along with a scraping system to fetch data from the dining site. After that we created a minimal front-end website, and then styled and improved it to become more user friendly and enjoyable. We then started feeding data to the back-end and tweaking the algorithm. Once the algorithm was mostly finished, we linked it to the Twilio API to send the SMS. Finally, we tested our creation and polish it more for the best user experience.

Challenges we ran into

It was difficult to obtain information about the food just from the name. For example, we want to be able to recommend the servery with mac and cheese if the user likes pasta. However, there are no ready-made API available for this, and we didn't have enough time to build a system of this complexity from scratch. We solved the problem by integrating Google Custom Search and Wikipedia's API to gain categorical information about the food.

Accomplishments that we're proud of

Even though we faced many challenges, we found ways to surmount them and managed to create something that is truly functional, practical, and useful for Rice students.

What we learned

All of us gained new experiences working with the technologies we used in this hackathon and familiarized ourselves with Node.js. We also learned how to manage our limited resources and get the most out of them.

What's next for Servery Hacks

  • Expand Servery Hacks to other college campuses.
  • Feed more information into our recommendation algorithm. For example, account for how far away each servery is from the user.
  • Improve our system of finding food information.
  • Make mobile apps for users who prefer mobile notifications over SMS.
Share this project: