Major project
Banking & Fintech
Stock movement prediction with news sentiment
Combine price history and news sentiment to study short-term moves of NIFTY 50 stocks.
The brief
An investment research desk wants to know whether news sentiment adds anything to price-based signals for Indian large-cap stocks. Build a sentiment model for financial news, combine it with price features, and test a simple strategy honestly (no peeking into the future). Good looks like: a sentiment model that beats a general-purpose one on financial text, a careful backtest with transaction costs, and a clear conclusion even if the answer is 'it doesn't help much'.
What you will build
- Prepare daily price features for the NIFTY 50 stocks
- Train and evaluate a financial-news sentiment classifier
- Join sentiment to prices without look-ahead bias
- Model next-day direction and compare with a price-only model
- Backtest a simple strategy with costs
- Report findings, risks and limits
Free, open data to build it with
- NIFTY-50 Stock Market Data (2000–2021) ↗ Kaggle
- Sentiment Analysis for Financial News ↗ Kaggle
- Financial PhraseBank ↗ Hugging Face
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.