EatConnect

User interface for EatConnect App

Inspiration

We first thought about having a platform that allows people to share videos and find friends based on similar interests. We then realized that a similar concept could be applied to food too! We could relate back to our previous experiences where we had difficulties thinking about a good place to eat at when we decided to eat out with our friends. This app could help recognize and recommend restaurants both to the individual user and to a group of user. We do not have much time now, but we plan to expand this further by making this like a social media platform that enables people who do not know each other previously to connect together based on their preferred categories. We are connecting people through food, one step at a time.

What it does

Recommends restaurants based on preferred food category learned through machine learning
Allows users to search for other users to match up preferences and provide recommendations suitable for both the user and the other user

How we built it

We split our task into 3 segments: Generation of machine learning data, development of machine learning model, and app development. We plan to use Firebase to integrate all three segments together so that first the machine learning data is used for the development of the supervised machine learning model. The user could input data through the app, then the data could be processed by the supervised machine learning model. The data is then retrieved and used to display a list of restaurant recommendations in the app.

The machine learning data is generated through an algorithm which randomizes input, with random 'preferred category' for each 'user' to mimic real-world app user behaviour. The machine learning model is then developed using Tensorflow and converted to Tensorflow Lite, to utilize Firebase's custom ML Kit for integration with the app. The app is developed using Android Studio.

Challenges we ran into

We did not have sufficient time. Our team only met up on the day of the hackathon and all three of us have varying interests and skill-sets: Java, web development and Machine Learning. We love the idea we had, but it involved a lot of learning for all of us and we spent way too much time setting up the environments we needed. We feel that if we were better prepared with all our environments set up before the hackathon, we could have done much more. :)

Accomplishments that we're proud of

THIS PROJECT! We love this idea and would probably move on with it even after the hackathon. We are also really proud because we have 2 people in EECS183 in our team but we've managed to actually code during the hackathon. We are proud of our team because our team works really well together.

What we learned

We gained new skill-sets, but at the same time, we also learned to better take care of our health and rest when we need to so that we can be more productive. We could also have done better if we've set up our environments prior to the hackathon.