What it does
A visualisation of messages travelling between countries, represented as navigating between nodes. Our aim is to demonstrate the potential conversations have as big data and the human values represented in them. We built a feature-rich, retro-themed web application: graph simulation, messenger and login page for a fun and interactive visualisation of people's questions and answers across cultural lines. We integrated Gemini API and Firestore for conversion and storage of messages as vector embeddings to find similarities and common themes among discussion.
Challenges we ran into
Building a web application with so many individual features proved to be difficult to maintain, especially within the hackathon 24 hours. The integration of Gemini API with Firestore was tough at times which made it feel so much better when everything fell into place.
Accomplishments that we're proud of
We're proud to have a working simulation, feature-full web application and AI ready within the hackathon time frame. And the challenges with technologies we have never touched before felt really rewarding. We learnt how to use Firebase for cloud storage for fast, real-time retrieval of messages across any device.
What's next for Space Mouse
- A top 3 ranking for local cuisines (requiring text-classification transformer models) and a ranking for countries based on number of inbound messages
- Particle simulation for better and more engaging visualisation of messages travelling between nodes.
Log in or sign up for Devpost to join the conversation.