Fiji News Intelligence Agent

A Windows application for harvesting, categorizing, and analyzing news about Fiji.

Features

• News Harvesting: Automatically collects news articles from major Fiji news sources
• News Categorization: Classifies articles into politics, community, sports, crime, and other categories
• Summary Generation: Creates comprehensive summaries of collected news
• Trend Analysis: Identifies emerging trends, potential threats, and suggests mitigation strategies
• Audio Conversion: Converts news summaries into speech for easy consumption

Installation

Prerequisites

• Python 3.8 or higher
• Windows 10 operating system

Setup

1. Clone or download this repository:

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

2. Create a virtual environment (recommended):

   python -m venv venv
   venv\Scripts\activate
   

3. Install the required dependencies:

   pip install -r requirements.txt
   

4. Create necessary directories:

   mkdir -p data
   

Usage

1. Start the application:

   python app.py
   

2. Open your web browser and navigate to:

   http://localhost:5000
   

3. Use the application:

   • Click "Harvest News" to collect the latest news about Fiji
   • View news articles categorized by topic
   • Generate a summary of all collected news
   • Analyze trends and identify potential threats
   • Convert summaries to audio for listening

News Sources

The application collects news from the following Fiji sources:

• Fiji Times
• Fiji Sun
• Fiji Village
• FBC News
• Islands Business

Technical Details

• Backend: Python with Flask web framework
• Frontend: HTML, CSS, JavaScript with Bootstrap 5
• News Harvesting: Newspaper3k and BeautifulSoup4
• Text Analysis: NLTK and scikit-learn
• Text-to-Speech: pyttsx3 (offline) and gTTS (online)

Customization

You can customize the application by:

1. Adding News Sources: Edit the sources list in news_harvester.py
2. Modifying Categories: Update the categories list and category_keywords in news_classifier.py
3. Enhancing Analysis: Extend the analysis capabilities in news_analyzer.py

Data Storage

The application stores data in the following formats:

• Harvested news: JSON files in the data directory
• Summaries: Text files in the data directory
• Analyses: JSON files in the data directory
• Audio files: MP3 files in the data directory

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Newspaper3k for article extraction
• NLTK for natural language processing
• Flask for the web framework
• Bootstrap for the UI components