POPNEWS is an intelligent news analysis system that combines state-of-the-art AI technology with structured workflows to provide comprehensive analysis of news articles. The system helps users
POPNEWS: AI-Powered News Analysis System
POPNEWS is an intelligent news analysis system that combines state-of-the-art AI technology with structured workflows to provide comprehensive analysis of news articles. The system helps users identify misinformation, understand article credibility, and gain key insights from news content in an increasingly complex information landscape.
POPNEWS offers a complete suite of analysis tools that evaluate news content from multiple perspectives:
Fake News Detection: Quantitative scoring (0-1) of articles with detailed reasoning
Content Summarization: AI-generated bullet points highlighting key insights
Fact Checking: Comprehensive credibility assessment with source verification
Topic-based News Search: Retrieval and analysis of multiple articles on a given topic
The system implements a sophisticated LangGraph-based agent that processes news content through a sequential workflow:
Orchestrated Workflow: Five-stage process ensures thorough analysis
Error Resilience: Robust error handling maintains functionality despite API failures
Modular Design: Each component can operate independently or as part of the pipeline
POPNEWS leverages Google's Gemini AI for complex language tasks:
Content Understanding: Deep semantic analysis of news articles
Critical Evaluation: Assessment of article credibility, bias, and accuracy
Information Extraction: Identification of key claims and insights
The heart of POPNEWS is its LangGraph-based agent workflow:
1. fetch_news()
Connects to GNews API to retrieve relevant articles
Transforms raw API responses into structured article objects
Implements search term encoding and response validation
Handles API rate limiting and error conditions gracefully
2. detect_fake_news()
Utilizes Gemini AI to analyze article content for misinformation
Generates a quantitative "fake news score" between 0 and 1
Provides detailed explanations for the assigned score
Identifies specific claims or content that raise concerns
3. generate_summary()
Creates concise, structured summaries of lengthy articles
Produces three focused bullet points per article covering:
Main news development
Important context/background
Potential impact or significance
Ensures consistency in output despite varying article quality
4. fact_check()
Performs comprehensive credibility assessment on a 1-10 scale
Evaluates source reputation and information quality
Identifies logical fallacies, misleading framing, or emotional manipulation
Suggests alternative trusted sources for verification
Provides specific warnings about potential misinformation
5. finalize_results()
Compiles all analysis data into a unified response structure
Ensures completeness and consistency across analysis components
Handles edge cases and missing information gracefully
POPNEWS exposes its capabilities through a RESTful API built with Flask:
1. GET /api/news
Searches and analyzes multiple news articles on a specified topic
Triggers the complete agent workflow from fetch to finalization
Returns fully analyzed articles with all metrics and insights
2. POST /api/analyze
Performs fact checking on a single user-provided article
Returns detailed credibility assessment and reliability points
Ideal for analyzing content from non-standard sources
3. POST /api/summary
Generates concise bullet-point summaries for user-provided articles
Extracts the most important information from lengthy content
Presents information in an easily digestible format
4. POST /api/fake-news-detection
Evaluates likelihood that an article contains misinformation
Provides a score and detailed explanation
Helps users quickly assess content reliability
The system integrates with external services for content and analysis:
1. GNews API
Provides up-to-date news articles from multiple sources
Supports topic-based searching and filtering
Returns structured article data including title, content, and source
2. Google Gemini AI API
Powers advanced natural language understanding and analysis
Enables sophisticated content evaluation and summarization
Provides reliable AI inference for detecting misinformation
The system uses a structured state object ( AgentState ) to maintain consistent data flow:
LangGraph enables a clear, sequential processing workflow:
Each analyzed article contains rich metadata and insights:
1. Comprehensive Analysis: Unlike tools that focus on a single aspect of news evaluation, POPNEWS provides multi-faceted analysis including fake news detection, fact checking, and summarization.
2. Agent-based Architecture: The LangGraph workflow enables reliable, consistent processing that can be extended with additional analysis stages.
3. Explanation-Focused: All evaluations include detailed explanations and reasoning, not just scores or ratings.
4. API-First Design: Easy integration with other applications and platforms through a clean REST API.
5. Error Resilience: Robust handling of API failures and edge cases ensures the system continues providing value even when external services are impaired.
1. User Feedback Integration: Incorporate user feedback to improve analysis quality and identify blind spots.
2. Expanded Source Database: Build an internal database of source credibility to reduce dependence on external AI.
3. Multilingual Support: Extend analysis capabilities to multiple languages.
4. Historical Trend Analysis: Track how narratives and reporting on topics evolve over time.
5. Browser Extension: Provide real-time analysis as users browse news sites.
Python 3.8+
Flask web framework
LangGraph for agent orchestration
GNews API access
Gemini API access (Google AI)
POPNEWS represents a significant advancement in automated news analysis, combining sophisticated AI capabilities with structured workflows to help users navigate today's complex information landscape. By providing detailed, multi-faceted analysis of news content, the system empowers users to make more informed decisions about the information they consume and share.
Everything was build during this hackathon .
NA