🧠 NewSense – Why Is My Fund Down?

NewSense is an AI-powered explanation system that helps investors understand why mutual funds or ETFs are falling by analyzing performance trends and scraping relevant real-world news. It uses NLP to summarize news articles, detect sentiment, highlight key entities, and match them with the fund's major holdings.

---

🚀 Features

• 🔍 Ticker Detection from natural language queries (e.g., "Why is QQQ down?")
• 📉 Fund Performance Visualization with Plotly
• 📰 Dynamic News Scraping based on ticker and keywords (no API dependency)
• 🧠 AI Summarization using NLP model
• 😃 Sentiment Analysis of scraped articles
• 🔎 Named Entity Recognition (NER) for companies, people, organizations, etc.
• 🎯 Matching News to Fund Holdings for better insights

---

🛠️ Tech Stack

• Python
• Streamlit (for the UI)
• Plotly (for charts)
• NLTK (for text processing)
• BeautifulSoup + Requests (for news scraping)
• Sentence-Transformers + FAISS (for semantic search)
• OpenAI / Transformers (optional: for summarization)

---

🖥️ Installation

1. Clone the repo

git clone https://github.com/yourusername/news-sense.git
cd news-sense

2. Create virtual environment (optional but recommended)

python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows

3. Install dependencies

pip install -r requirements.txt

4. Download NLTK data (optional: handled in app)

import nltk
nltk.download('punkt')
nltk.download('stopwords')

▶️ Running the App

streamlit run app.py

Then open in browser: http://localhost:8501

📁 Project Structure

.
├── app.py                  # Main Streamlit app
├── fetch_fund_data.py      # Pulls mutual fund data via Alpha Vantage API
├── news_scraper.py         # Custom Bing/Google news scraper
├── model_run.py            # NLP pipeline (summarization, sentiment, NER)
├── data/
│   ├── fund_data_<TICKER>.json
│   └── news/news_<TICKER>.json
├── requirements.txt
└── README.md

💡 Example Queries

"Why is QQQ falling?"

"What’s affecting SPY today?"

"News around SBI mutual fund"

"Why is the tech sector down?"

🙌 Credits

Built by Team NewsSense at the "Why Is My Fund Down?" Hackathon Challenge 🧠 Big thanks to open-source libraries and APIs used.

📄 License

MIT License – free to use, modify, and share.

---
Let me know if you want a version that includes screenshots, GitHub badges, or OpenAI API setup instructions.