Dakota AI Demo: Fake News Detection with NLP

# Dakota AI Demo: Fake News Detection with NLP
# Showcase: Data Processing & Analytics, Custom AI Solutions
# Dataset: Fake and Real News Dataset (https://kaggle.com/datasets/clmentbisaillon/fake-and-real-news-dataset)
# Run in Google Colab (Free): https://colab.research.google.com/

import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.naive_bayes import MultinomialNB
from sklearn.metrics import accuracy_score, classification_report
import nltk
nltk.download('punkt', quiet=True)
from wordcloud import WordCloud

print("Loading Fake News dataset...")
# Using sample data for demo (full dataset requires more processing time)
fake_data = pd.DataFrame({
    'title': ['BREAKING: New Evidence Found in UFO Case',
              'Government Confirms Alien Contact',
              'Secret Meetings Exposed in Leaks',
              'Climate Change is a Hoax Claims Scientist',
              'Election Was Rigged, Sources Say',
              'BREAKING: Massive Economy Crash Coming',
              'Health Study: Vaccines Cause Autism',
              'Foreign Power Bribed Officials',
              'Moon Landing Was Faked, New Evidence',
              'Doctors Warn of Deadly Poison in Food'],
    'label': 0
})

real_data = pd.DataFrame({
    'title': ['President to Address Nation on Climate Change',
              'New Study Confirms Benefits of Exercise',
              'Stock Market Reaches New Highs',
              'Scientists Discover New Planet',
              'Company Announces 20% Growth',
              'Local Fire Department Saves Family',
              'Schools Plan New Technology Program',
              'New Medical Treatment Shows Promise',
              'Weather Forecast: Clear Skies Ahead',
              'State Budget Increases Education Funding'],
    'label': 1
})

df = pd.concat([fake_data, real_data], ignore_index=True)
print(f"Dataset Shape: {df.shape}")
print(f"Fake News: {len(df[df['label'] == 0])} articles")
print(f"Real News: {len(df[df['label'] == 1])} articles")
print()

print("=== Data Processing & Analytics ===")
print("1. Text Preprocessing:")
df['processed_title'] = df['title'].apply(lambda x: x.lower().replace('breaking:', '').replace('claims', '').strip())
print("- Cleaned titles and prepared for vectorization")
print()

print("2. Feature Extraction:")
vectorizer = TfidfVectorizer(max_features=500, stop_words='english')
X = vectorizer.fit_transform(df['processed_title'])
y = df['label']
print(f"TF-IDF Matrix Shape: {X.shape}")
print(f"Top Features: {vectorizer.get_feature_names_out()[:5]}")
print()

print("=== Custom AI Solutions ===")
print("3. Model Training:")
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)

model = MultinomialNB()
model.fit(X_train, y_train)
y_pred = model.predict(X_test)

accuracy = accuracy_score(y_test, y_pred)
print(".2f")
print()
print("Classification Report:")
print(classification_report(y_test, y_pred, target_names=['Fake News', 'Real News']))
print()

print("4. Visualization - Word Clouds:")
fake_text = ' '.join(df[df['label'] == 0]['processed_title'])
real_text = ' '.join(df[df['label'] == 1]['processed_title'])

plt.figure(figsize=(12, 6))
plt.suptitle('Dakota AI: Fake vs Real News Text Analysis')

plt.subplot(1, 2, 1)
wordcloud_fake = WordCloud(width=400, height=300, background_color='white').generate(fake_text)
plt.imshow(wordcloud_fake, interpolation='bilinear')
plt.title('Fake News Word Patterns')
plt.axis('off')

plt.subplot(1, 2, 2)
wordcloud_real = WordCloud(width=400, height=300, background_color='white').generate(real_text)
plt.imshow(wordcloud_real, interpolation='bilinear')
plt.title('Real News Word Patterns')
plt.axis('off')

plt.tight_layout()
plt.savefig('fake_news_wordclouds.png', dpi=150)
print("Word clouds saved as 'fake_news_wordclouds.png'")
print()

print("=== Demo Summary ===")
print("✓ Processed news article text data")
print("✓ Applied TF-IDF vectorization for feature extraction")
print("✓ Trained Naive Bayes classifier for fake/real detection")
print("✓ Generated word cloud visualizations")
print(f"✓ Achieved {accuracy:.2f} accuracy on test set")
print()
print("This demonstrates Dakota AI's NLP and text classification expertise!")