machine learningbertyahoo finance
Sentiment Analysis
Published 5.3.2024
Overview
This project aimed to classify sentiment of financial news. Developed for my bachelor thesis, it showcases how financial news can impact market sentiment.
Features
- Input Yahoo Finance News articles to be analysed
- See real-time status of the analysis
- View the sentiment of each sentence with a overall sentiment score
Tech Stack
- Frontend: [Next.js, Typescript, TailwindCSS, shadcn/ui]
- Backend: [AWS Lambda, AWS API Gateway, Python, BERT (Huggingface)]
- Database: [SQLite, PostgreSQL]
Challenges and Solutions
One of the main challenges faced during this project was creating my own dataset and training a pre-trained BERT model with it. Another challenge was to create a real-time analysis system, which was solved by using AWS Lambda and API Gateway.
Results and Impact
The project was successfully launched in January 2024 and has been praised by my professors and peers.
Lessons Learned
Throughout this project, I learned how to create datasets, train a BERT model, and deploy a real-time analysis system. I also had the opportunity to work with AWS Lambda and API Gateway, which was a great learning experience.