Happy Map

Flowchart of our project
HappyMaps Website

Inspiration

We wanted to go beyond just visualizing places—what if we could visualize how people feel in those places? With rising global uncertainty, we felt it was important to build something that reflects the emotional state of communities using real, live data.

What it does

HappyMaps collects real-time data from public news and Reddit APIs, analyzes the emotional tone of each post using NLP and backpropagation, and calculates a happiness score for each city. These scores are then displayed on a vibrant, interactive heatmap powered by Google Maps.

How we built it

We wrote Python scripts to fetch and preprocess live data from Reddit and news APIs. We used NLP techniques for sentiment analysis and implemented a weighted scoring system refined via backpropagation. The scores were stored in MongoDB, accessed by a Node.js backend, and finally visualized through the Google Maps API on our frontend.

Challenges we ran into

Handling noisy and irrelevant data from public APIs

Fine-tuning the scoring system to reflect real emotional tone

Making NLP work efficiently with limited time and mixed sources

Integrating multiple tech stacks (Python, Node.js, MongoDB, Google Maps) smoothly

Accomplishments that we're proud of

Successfully building a functional emotional heatmap from scratch

Creating a scoring algorithm that balances multiple sources fairly

Seamlessly connecting data processing, backend, and frontend

Making the idea of “mapping emotions” actually work and feel intuitive

What we learned

The power (and complexity) of combining NLP with real-world data

How to structure a full-stack project from API ingestion to interactive visualization

The importance of filtering and weighting in emotion-based analysis

How to collaborate and adapt quickly when dealing with multiple tech tools

What's next for Happy Map

Expanding to more locations

Expanding to include more data sources like Twitter or weather patterns

Adding time-based trends to track mood shifts over days or weeks

Introducing user filters (e.g., age groups, topics) for personalized views

Training a custom sentiment model for better context-aware scoring

Turning HappyMap into a tool for cities, researchers, or travelers to understand emotional geography

Built With

Submitted to

AI Hackfest

Created by

I led the data pipeline for HappyMap — writing Python scripts to collect real-time data from Reddit and news APIs, cleaning and preprocessing the text, and implementing the NLP-based sentiment analysis system. I also helped design the custom scoring logic using weighted averages and backpropagation to generate accurate happiness scores.

Shlok singh
I was responsible for building the frontend of the project using HTML, CSS, JavaScript, and React. I designed and developed the interactive map interface using the Google Maps API, ensuring real-time updates of emotion data markers based on user location. I also integrated the API responses with the map and handled dynamic rendering of sentiment-based markers to create an engaging and visually informative experience. . I also worked on integrating data from the backend into the map view, enabling seamless updates and smooth user interaction with features like info windows, location-based filtering, and intuitive visuals.

Revanth Anil Kumar
I worked on integrating the Twitter API to fetch relevant tweets. Although this feature wasn't included in the MVP due to scalability issues and limited data generation, I plan to scale it further and integrate Twitter data into our HappyMap in the future. I also modified existing Python scripts to address scaling challenges and handled the configuration of API keys and .env files. Additionally, I contributed to the UI and design aspects of the project.

Kaustubh Dwivedi
My role was of a Database Engineer and Manager, as well as the Backend Engineer, my work included collaborating with everyone, to put the fetched data into the database, using the database to retrieve data on command to help in the calculation of the score and then put the final data in a collection named 'emotion' in our case.

I also created the backend of our project, in which I created an API for the front-end to fetch the data from, the API I created accessed the 'emotion' collection in our database, and retrieved all the data from the same, which was fetched by the frontend and used to complete our project.

Piyush Rathi