With the magnificent launches of SpaceX and NASA’s Perseverance rover, many are not aware of such amazing breakthroughs in science and as a team of space enthusiasts, we are always wanting to know more about the recent space discoveries and the latest news. Although there are tonnes of resources available on the internet, most of the news feeds are not customizable and are not tailored towards one’s specific interests. In light of that issue, we decided to create a website that provides a space-related news feed, integrated with a machine learning algorithm to recommend articles the user is interested in, and provide a great experience for the user.

What it does

Rocket News provides a one-touch space-related news feed based on the latest articles on the web, and the user’s preferences. Users can stay on top of the latest space related news and recommended articles using our platform.

How we built it

We used Figma to design the website and the UI/UX. The actual frontend was created using technologies like HTML, CSS and Javascript. The backend was built using the Python framework Flask. We used the Newscatcher api to build the news feed, which incorporates pagination using the Rest API we created. Recommended articles are generated using data analytics and the machine learning algorithm. We used Scikit Learn, a Python Library, to build the algorithm for recommendations. The database was created using Google’s Firebase Database on the cloud. The application is deployed to Google Cloud using Google App Engine which takes care of all the infrastructure related issues.

Challenges we ran into

One of the main problems for us was implementing the pagination and Rest API. This was a new concept for us and something we had never worked with before. When working with the recommendation algorithm, we ran into issues surrounding data piping. To combat this, we examined the documentation thoroughly and found a fix. We also encountered a few problems with deploying to Google Cloud and dealing with the Firebase Database, and learnt a lot by attempting to solve them.

Accomplishments that we're proud of

We are proud to ship a full fledged website on the cloud available to all users with basic internet access. The website is integrated with a login system, a fully functional database, a Rest API for dynamically fetching data, and a high level recommendation algorithm based on user data. We are also proud of setting ourselves apart from other recommender systems by using support vector machines. We are very proud of what we did in the short span of time that we had.

What we learned

We learned about a variety of technologies and concepts. This was our first time implementing a Rest API with pagination and we learnt a lot from the process. Our website includes several features that are part of modern web applications, and building it helped us get a better understanding of how technologies work together. Because we had many problems and had to come up with several solutions, we gained a lot of experience and insight in a very short span of time. We also enhanced our interpersonal skills working in a team setting as we’ve never worked together. This hackathon was a great learning experience for all of us.

What's next for Rocket Class

Going ahead, we would like to develop a community platform that lets users communicate with others and have fascinating discussions about space and other related topics. In the future, we hope to make our product recommendation more thorough. We also hope to make our product recommend space related courses as well as news articles, so the user is able to learn more from our site. We are also planning to improve our machine learning model by using different ways of sorting data, such as k-means clustering or using different sigmoid functions to train and test our model.

New Features

1. Community building

Encourage community building by adding features such as friends, private messages, forums, and discussion threads.

2. Gamification component

Make the interaction process fun and enjoyable by adding features such as leader board, arena and battleground with space quizzes.

Share this project: