EZQuiz

Inspiration

We were inspired by the idea of Kahoot and how it's a very fun but also competitive quiz-type game. The main issue was that you could easily guess or use the process of elimination to get points, even if you don't completely know the answer.

What it does

EZQuiz allows someone to create a room where they can host a quiz, and any number of people can join based on the auto-generated code. The people who join then get the chance to add questions to the question pool and the answers to the questions. Then, when the quiz starts, everyone will get to answer the questions in text and not multiple choice. A natural language processor encodes the sentences. We use cosine similarity to compare the answer to the person who wrote the code's answer and give the players their points based on their similarity to the original answer.

How we built it

To build this, we first used node.js to set up our entire site. Then, we used Github to control our version and back up our documents in case of any losses. Using express as our backend and react.js as our front end, we successfully integrated the two to create a smooth user experience. Most importantly, we used HTML and CSS - like the tag - to introduce animations into the game to create more fun. Lastly, we have implemented TensorFlow(AI) to use as a "judge" for our game, using a pre-trained model to determine how similar two answers are and reward by how accurate it is.

Challenges we ran into

We could not publish the website to anything like GitHub pages because we used react.js instead of raw HTML. We also could not host the game's backend because there are no free Node.js hosting services.

Accomplishments that we're proud of

We created the entire website, and people can play together and learn. The app also implements thousands of lines of code and many complex (elaborate) logic that all come together to form what we know as EZQuiz.

What we learned

We learned a lot more about syntax and how to create websites in React, as well as about NLPs and how TensorFlow compares sentences. Additionally, we learned the complexity of the backend, with many problems connecting the backend with the front end and the server side. This was especially hard when we tried to put our code onto a runnable, global website when we got to connect to the GoDaddy site, which we learned was pretty hard.

What's next for EZQuiz

Next up, we will undoubtedly expand our website to a global market, where we could implement our code onto our GoDaddy server and prepare for the global launch of EZQuiz. Then, we are likely to add more modes to the game, introducing more tension and competition elements that we could not make in the 24 hours we have. Additionally, we will undoubtedly employ more aspects of art, including music and artwork.