TrendPulse

Real-Time Sentiment, Smarter Insights.

Comment

Inspiration

When people participate in stock market, it's always important to focus on social sentiment. With chatter on social platforms reflecting markets more than ever, we were inspired to create a tool that translates this vast, real-time data into actionable insights for smarter financial decision-making.

What it does

TrendPulse scans multiple social media channels for relevant keywords related to user's preferences for stocks and cryptocurrencies. By leveraging advanced NLP models, it analyzes posts for sentiment, aggregating data to produce a real-time market trend and risk index that helps users gauge market mood and make informed decisions.

How we built it

Data Ingestion: We integrated APIs from platforms like Twitter and Reddit, streaming data via Apache Kafka for real-time processing.

NLP Analysis: Utilizing state-of-the-art NLP to extract sentiment and classify keywords.

Backend Services: Built with Python to process data and aggregate sentiment scores, stored securely in MongoDB

Frontend: Developed a responsive dashboard with React.js to visualize trends and alerts, supported by real-time updates through WebSockets.

Challenges we ran into

Harmonizing data across multiple platforms with varying API structure. Ensuring smooth data flow from backend to frontend was initially tricky. We faced issues where the backend worked fine in isolation, but the frontend couldn’t render the data properly due to async handling and data structure mismatches. Due to time constraint we use some mock data in this demo.

Accomplishments that we're proud of

Having real-time market data and corresponding social media data, along with the sentimental analysis and score.

What we learned

How to use social media API to get data, how to make backend API endpoints and connect with React front end

What's next for TrendPulse

Due to the time constraint, there are so many things we haven't done - like on a cloud server, Integrate Twitter/X API for broader social media coverage, Add financial news API integration (Bloomberg, Reuters, etc.), Implement Discord and Telegram community sentiment analysis For Analytics part, we're looking for: Develop predictive models to forecast price movements based on sentiment Create correlation analysis between sentiment shifts and price action Implement machine learning algorithms to improve sentiment accuracy Add historical trend analysis with advanced visualizations Develop a custom sentiment scoring system that weights social media sources by influence

Share this project:

Updates