🏏 Cricket Fact Checker (AI-Assisted)

An AI-assisted, editorial-grade cricket fact-checking application built using Python and Streamlit, designed for cricket writers, editors, and content teams.

The app verifies cricket facts using trusted public sources:

• ESPNcricinfo
• Cricbuzz
• Sportskeeda

It supports AI-generated articles, Word documents with tables, and PDF files, and shows ONLY incorrect facts — no noise, no confusing scores, no unnecessary warnings.

---

🎯 Problem This App Solves

Cricket content often contains:

• Incorrect stats in AI-generated or human-written articles
• Errors hidden inside Word tables
• Time-consuming manual fact-checking
• False alarms from over-aggressive tools

This app is built with an editor-first mindset:

• It flags only clearly incorrect facts
• It stays silent when unsure
• It avoids false positives
• It supports tables inside Word files

---

✨ Key Features

✅ Writer-First Fact Checking

• Shows ONLY incorrect cricket facts
• No accuracy percentages or confusing labels
• Silent when facts are true or unverifiable
• Editorially safe behavior

🧠 Open-Source AI Integration

• Uses sentence-transformers/all-MPNet-base-v2
• Understands:
  • AI-generated articles (ChatGPT, Claude, etc.)
  • Humanized AI content
  • Long, polished editorial writing
• AI is used only for language understanding
• AI never invents or decides facts

📄 Multiple Input Methods

• Paste full article text
• Upload Word (.docx) files
• Upload PDF (.pdf) files

📊 Word Table Fact Checking (Major Highlight)

• Automatically reads tables inside Word files
• Converts each table row into factual sentences
• Verifies table data using trusted sources
• Identifies exactly:
  • Table number
  • Row number
  • Column name
• Displays table errors clearly in UI and reports

📌 Clear Output

• ❌ Lists only incorrect facts
• ✅ Shows “Safe to publish” when no issues are found
• No “Needs fact check” or yellow warnings

📥 Exportable Reports

• Download PDF report (incorrect facts only)
• Download Word (.docx) report
• Table errors included with row & column references

🎨 Clean UI & Animations

• Loading spinner during processing
• Progress bar while checking
• Simple, editor-friendly interface

---

🧠 How the App Works (Simple Flow)

---

🧱 Tech Stack

• Python 3.9+
• Streamlit (UI)
• spaCy (sentence processing)
• Sentence Transformers – MPNet (AI understanding)
• Requests + BeautifulSoup (source scraping)
• python-docx (Word & table parsing)
• PyPDF2 (PDF text extraction)
• ReportLab (PDF report generation)

---

📁 Project Structure

cricket_fact_checker/ │ ├── app.py # Main Streamlit app ├── ai_helper.py # Open-source AI (MPNet) ├── utils.py # Text utilities ├── verifier.py # Fact verification logic ├── exporter.py # PDF & Word report export ├── file_reader.py # Word/PDF + table reader │ └── scrapers/ ├── init.py ├── espn.py ├── cricbuzz.py └── sportskeeda.py

---

🚀 How to Run Locally

1️⃣ Clone the Repository

git clone https://github.com/yourusername/cricket-fact-checker.git
cd cricket-fact-checker
2️⃣ Create Virtual Environment
python -m venv venv
venv\Scripts\activate   # Windows
# source venv/bin/activate  # macOS/Linux

3️⃣ Install Dependencies
pip install -r requirements.txt

4️⃣ Download spaCy Model
python -m spacy download en_core_web_sm

5️⃣ Run the App
streamlit run app.py


Open in browser:

http://localhost:8501