CryptoMeter

Inspiration

Fairly recently we saw a story about a guy who created a bot to buy dogecoin every time Elon Musk tweeted. It got us thinking about how much public perception on social media matters to the price of cryptocurrencies. Could a single tweet cause a crypto's price to skyrocket or plummet. After many hours of model testing we created CryptoMeter.

What it does

CryptoMeter scrapes tweets relating to Bitcoin and Ethereum then calculates the sentiment of those tweets. Using the sentiment scores we created a model that will predict the prices of Bitcoin and Ethereum.

How we built it

We built the backend through python using pandas, sklearn, snscrape, transformers, scipy. transformers and scipy are part of the Roberta model that we used for sentiment analysis. sklearn allowed us to make a linear regression model for this project and snscrape allowed us to get all the tweets we needed.

Challenges we ran into

One of the major challenges we ran into was making the model. Mostly because we had to comb through a lot of the data and fine tune it for it to be able to make it a usable model. We faced a challenge when designing a webpage using CSS. However, we were able to overcome this challenge by utilizing the power of flexbox and SCSS. Flexbox is a layout module in CSS that makes it easy to align and distribute elements within a container. By using the flexbox properties, we were able to easily align elements both horizontally and vertically, and control the size and spacing of elements. Additionally, by using SCSS, we were able to write more efficient and organized CSS code. SCSS allowed us to use variables, nested selectors, and other advanced features which made the process of writing CSS much more manageable. Overall, using flexbox and SCSS was a game-changer for us in terms of webpage design and we highly recommend them to anyone facing similar challenges.