What is Collaborative Filtering?

Collaborative Filtering (CF) is a recommendation technique that predicts what a user might like based on the preferences and behavior of other similar users.

Why Collaborative Filtering is Important?

• Powers most modern recommendation engines.
• Works without needing item content data.
• Suggests new products outside your usual interests.
• Improves customer engagement and retention.
• Used by Netflix, Amazon, Spotify, YouTube.

Real-Life Examples

• Netflix recommends movies that users similar to you watched.
• Amazon suggests products users like you purchased.
• Spotify creates playlists based on similar users.
• YouTube shows videos based on similar viewers.

How Collaborative Filtering Works

Types of Collaborative Filtering

Mathematical Formula

Predicted rating using user-based CF:

Where:

• r̂ = predicted rating
• u, v = users
• i = item
• sim(u,v) = similarity score

Similarity Metrics

Example: User-Item Matrix

Python Example — Item-Based CF

pip install pandas scikit-learn

import pandas as pd
from sklearn.metrics.pairwise import cosine_similarity

# Sample User-Item Matrix
data = pd.DataFrame({
    "Movie A": [5,4,0,3],
    "Movie B": [4,0,5,0],
    "Movie C": [0,5,4,0],
    "Movie D": [2,0,1,5]
}, index=["U1","U2","U3","U4"])

# Compute item similarity
similarity = cosine_similarity(data.T)
print(pd.DataFrame(similarity, index=data.columns, columns=data.columns))

Real-Life Analogy

Visual Workflow

Real-World Applications

Advantages

• No need for item content data.
• Suggests new items beyond user history.
• Captures real-time behavior patterns.
• Works for big platforms.
• Highly personalized recommendations.

Disadvantages

Common Mistakes

Best Practices

Importance of Collaborative Filtering

Golden Rule