flsteven87/InfoEssence
Steven Wu
GlobalNews for Taiwan
GlobalNews for Taiwan is an advanced news aggregation and summarization system designed to provide Taiwanese readers with concise, relevant international news in Traditional Chinese. This project leverages AI technologies to collect, summarize, and present global news in a format tailored for the Taiwanese audience.
Features
News Aggregation: Collects news from various international sources using RSS feeds.
AI-Powered Summarization: Utilizes OpenAI's GPT models to generate concise summaries and titles in Traditional Chinese.
Important News Selection: Automatically selects the most relevant and important news items for the Taiwanese audience.
Instagram Post Generation: Creates engaging Instagram-ready posts from selected news items.
Image Generation: Generates relevant images for news articles using AI.
Web Interface: Provides a user-friendly web interface for browsing summarized news.
Technology Stack
Backend: Python
Database: PostgreSQL
ORM: SQLAlchemy
API: OpenAI GPT-4
Web Framework: Streamlit
Image Generation: DALL-E (via OpenAI API)
Setup and Installation
Clone the repository:
Install dependencies:
Set up environment variables: Create a
.env file in the root directory and add the following:Initialize the database:
Run the application:
Usage
The main script supports various operations that can be executed as modules.
Updating RSS Feeds
To update the RSS feeds:
Fetching Latest News
To fetch and process the latest news articles:
Selecting Important News
To select a specific number of important news items (e.g., 5):
This will select the 5 most important news items and save them to a CSV file in the
chosen_news directory.Generating Instagram Posts
Instagram posts are automatically generated for the chosen news items and saved to a CSV file in the
instagram_posts directory.Running the Web Interface
To start the Streamlit web interface:
Typical Workflow
A typical workflow might look like this:
Update RSS feeds
Fetch latest news
Choose important news items
View results in the web interface
Project Structure
src/: Contains the main source codeconfig/: Configuration files (config.yaml, rss_feed.yaml)database/: Database models and operationsservices/: Core services for news processingutils/: Utility functionsrequirements.txt: List of Python dependencies.gitignore: Specifies intentionally untracked files to ignoreLicense
MIT
Acknowledgements
OpenAI for providing the GPT models used in summarization
All the news sources that provide RSS feeds
Like this project
Posted Mar 30, 2025
Contribute to flsteven87/InfoEssence development by creating an account on GitHub.
Likes
0
Views
0
