🤖 AI Web Scraping Tool with Fact-Checking

Powered by DeepSeek V3 & DeepSeek R1 via OpenRouter

Extract and analyze web content using advanced AI, then fact-check the results for accuracy and reliability.

🔄 How to Use the Fact-Checking Feature:

  1. First: Enter your API key, URL, and analysis query
  2. Second: Click "🚀 Analyze Website" to get initial results
  3. Third: Click "🔍 Fact-Check Results" to verify accuracy with DeepSeek R1

🎯 What the Fact-Checker Does:

Accuracy Verification

  • Compares every claim in the analysis against the original source
  • Identifies factual errors and misrepresentations
  • Verifies numerical data and statistics

Completeness Assessment

  • Checks if important information was missed
  • Evaluates coverage of all relevant aspects
  • Identifies gaps in the analysis

Context Verification

  • Ensures information isn't taken out of context
  • Verifies proper interpretation of source material
  • Checks for misleading presentations

Quality Scoring

  • Provides accuracy scores (1-10 scale)
  • Lists verified vs. unverified claims
  • Offers specific recommendations for improvement

🧪 Best Practices for Fact-Checking:

Ideal Test Cases:

URL: https://en.wikipedia.org/wiki/List_of_countries_by_population
Query: Create a table showing the top 10 most populous countries with their exact population figures

Perfect for fact-checking numerical accuracy

URL: https://www.who.int/news-room/fact-sheets
Query: Extract key health statistics and create a summary of global health metrics

Great for verifying official statistics

URL: https://finance.yahoo.com/quote/AAPL
Query: Extract Apple's current stock price, market cap, and financial metrics

Excellent for checking real-time financial data accuracy

🎯 Example Analysis Queries for Fact-Checking:

Data-Heavy Content

  • "Extract all numerical data and organize it in a table format"
  • "Create a comparison table of different countries' GDP figures"
  • "List the top 10 items with their exact values from the source"

Statistical Information

  • "Summarize key statistics with specific numbers and percentages"
  • "Extract survey results and present the exact figures"
  • "Create a timeline with specific dates and events"

Complex Analysis

  • "Compare different viewpoints and cite specific quotes"
  • "Extract cause-and-effect relationships mentioned in the article"
  • "Summarize research findings with methodology details"

🔍 What Gets Fact-Checked:

✅ Verified Items:

  • Exact quotes and citations
  • Numerical data and statistics
  • Dates, names, and factual claims
  • Table data accuracy
  • Mathematical calculations

⚠️ Flagged Issues:

  • Misquoted information
  • Incorrect numbers or percentages
  • Missing context or nuance
  • Overgeneralized statements
  • Unsupported conclusions

🚨 Red Flags the Fact-Checker Catches:

  • Hallucinated Data: Information not present in the source
  • Misattributed Quotes: Quotes assigned to wrong sources
  • Mathematical Errors: Incorrect calculations or summaries
  • Context Loss: Information presented without proper context
  • Incomplete Extraction: Missing important details from tables

💡 Tips for Better Fact-Checking:

  1. Use Specific Queries: More specific requests = better fact-checking
  2. Test with Known Data: Start with sites where you know the content
  3. Check Complex Tables: Tables are great for testing accuracy
  4. Verify Names & Dates: These are common error points
  5. Cross-Reference: Compare with multiple sources when possible

🔬 Advanced Fact-Checking Tests:

Financial Data Test

URL: https://finance.yahoo.com/quote/MSFT
Query: Create a detailed financial summary table with exact figures for Microsoft stock
Expected: Fact-checker should verify all numbers match the source exactly

Statistical Data Test

URL: https://www.census.gov/quickfacts/fact/table/US
Query: Extract US population demographics with specific percentages
Expected: Fact-checker should confirm all demographic percentages are accurate

Historical Data Test

URL: https://en.wikipedia.org/wiki/List_of_Presidents_of_the_United_States
Query: Create a table of the last 10 US presidents with their exact terms of office
Expected: Fact-checker should verify all dates and names are correct

🧪 Test Scenarios

1. News & Media Sites

URL: https://www.bbc.com/news
Query: Extract the top 5 news headlines with their summaries and create a table with columns: Headline, Category, Summary

URL: https://edition.cnn.com
Query: Find all breaking news items and organize them by topic/region in a structured format

2. Financial Data Sites

URL: https://finance.yahoo.com/quote/AAPL
Query: Extract Apple stock information including current price, daily change, market cap, and any financial metrics into a summary table

URL: https://www.marketwatch.com/investing/stock/tsla
Query: Create a table with Tesla's key financial metrics: price, change, volume, market cap, P/E ratio

3. E-commerce & Product Pages

URL: https://www.amazon.com/dp/B08N5WRWNW
Query: Extract product details including name, price, ratings, key features, and specifications in a structured format

URL: https://www.ebay.com/itm/123456789
Query: Extract item details, price, seller information, and shipping details into a comparison-ready table

4. Educational & Reference Sites

URL: https://en.wikipedia.org/wiki/Artificial_intelligence
Query: Extract the main definition, history timeline, and applications of AI. Create separate sections for each topic.

URL: https://en.wikipedia.org/wiki/List_of_countries_by_population
Query: Extract the population data table and create a new table showing top 10 most populous countries with their population and growth rate

5. Government & Official Statistics

URL: https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports
Query: Extract the latest COVID-19 statistics and create a summary table with key global figures

URL: https://www.census.gov/quickfacts
Query: Extract key demographic statistics for the United States and organize them into categories: Population, Economy, Geography

6. Technology & Business News

URL: https://techcrunch.com
Query: Find the latest startup funding news and create a table with: Company Name, Funding Amount, Investors, Industry

URL: https://www.reuters.com/technology
Query: Extract top technology news and summarize each story in 2-3 sentences with key points

7. Scientific & Research Sites

URL: https://www.nature.com/articles
Query: Extract recent scientific article titles, authors, and abstracts. Create a summary table organized by research field

URL: https://pubmed.ncbi.nlm.nih.gov/trending
Query: Find trending medical research topics and create a list with brief descriptions of each study's findings

8. Sports & Entertainment

URL: https://www.espn.com/nba/standings
Query: Extract NBA team standings and create a table with: Team, Wins, Losses, Win Percentage, Conference Position

URL: https://www.imdb.com/chart/top
Query: Extract the top 10 movies from IMDb's top 250 list with ratings, year, and brief description

9. Weather & Environmental Data

URL: https://weather.com/weather/today
Query: Extract current weather conditions and forecast data. Create a summary with temperature, conditions, and weekly outlook

10. Real Estate & Property

URL: https://www.zillow.com/homes/for_sale
Query: Extract property listings with prices, locations, square footage, and key features into a comparison table

🎯 Quick Test Samples (Copy & Paste Ready)

Simple Test:

URL: https://httpbin.org/html
Query: Extract all text content and identify the page structure

Table Extraction Test:

URL: https://www.w3schools.com/html/html_tables.asp
Query: Find all HTML tables on this page and convert them to a structured format with proper headers

Complex Analysis Test:

URL: https://www.sec.gov/edgar/browse/?CIK=320193
Query: Extract Apple Inc.'s recent SEC filings and create a table with: Filing Date, Document Type, Description

International Site Test:

URL: https://www.bbc.co.uk/weather
Query: Extract UK weather information and create a regional breakdown of current conditions

🎯 Interpreting Fact-Check Results:

Accuracy Scores:

  • 9-10: Highly accurate, minimal issues
  • 7-8: Generally accurate with minor corrections needed
  • 5-6: Moderate accuracy, several issues to address
  • 3-4: Low accuracy, significant problems found
  • 1-2: Poor accuracy, major fact-checking failures

Verification Status:

  • ✅ VERIFIED: Claim matches source exactly
  • ⚠️ PARTIALLY VERIFIED: Claim is mostly correct but lacks nuance
  • ❌ CANNOT VERIFY: Claim not supported by source material
  • 🚨 CONTRADICTED: Claim directly contradicts source

Remember: The fact-checker is designed to be thorough and critical. Even high-quality analyses may receive suggestions for improvement!