≈ IT Science and Life Seminars

# Simple recommendation algorithm
def recommend_content(user_interests, mood):
    if "technology" in user_interests and mood == "curious":
        return "Latest AI research articles"
    elif "cooking" in user_interests:
        return "New recipe tutorials"
    else:
        return "Trending topics"

# How might this algorithm create a filter bubble?

# Simple data analysis example
import pandas as pd
import matplotlib.pyplot as plt

# Load and explore data
data = pd.read_csv('social_media_usage.csv')
print(data.describe())

# Visualize time spent by platform
platforms = data['platform']
time_spent = data['minutes_per_day']

plt.bar(platforms, time_spent)
plt.title('Daily Time Spent on Social Media')
plt.xticks(rotation=45)
plt.show()

# Simple web scraping example (ethical use only)
import requests
from bs4 import BeautifulSoup

# Get public information (always check robots.txt and terms of service)
url = 'https://example.com/public-data'
response = requests.get(url)
soup = BeautifulSoup(response.content, 'html.parser')

# Extract specific information
titles = soup.find_all('h2')
for title in titles:
    print(title.get_text())

# Discussion: How does web scraping relate to privacy?

# Simple AI example with pre-trained model
from transformers import pipeline

# Use a simple sentiment analysis model
classifier = pipeline('sentiment-analysis')

# Analyze text sentiment
results = classifier([
    "I love this course!",
    "I'm frustrated with this technology.",
    "The weather is neutral today."
])

for result in results:
    print(f"Text: {result['label']} with confidence: {result['score']:.2f}")

# Discussion: What patterns is the model detecting?

# Simple linear regression example
import numpy as np
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression

# Sample data: house sizes and prices
size = np.array([50, 60, 70, 80, 90, 100]).reshape(-1, 1)
price = np.array([150000, 180000, 200000, 220000, 240000, 260000])

# Create and train model
model = LinearRegression()
model.fit(size, price)

# Make prediction
prediction = model.predict([[85]])
print(f"Predicted price for 85m² house: €{prediction[0]:,.2f}")

# Discussion: What assumptions does this model make?

# Concept of model interpretation (simplified)
# In practice, you would use libraries like LIME or SHAP

def explain_prediction(model, input_data, feature_names):
    # Simplified explanation generator
    coefficients = model.coef_
    
    print("Feature contributions to prediction:")
    for i, feature in enumerate(feature_names):
        contribution = coefficients[i] * input_data[i]
        print(f"  {feature}: {contribution:.2f}")
    
    return "Transparent decision-making enables accountability"

# Discussion: When is transparency more important than accuracy?