อัลกอริทึมวัดความเป็นศูนย์กลาง (PageRank)

In a network, some nodes are more "important" than others. But how do we define importance?

Centrality algorithms help us measure a node's significance within a graph. They reveal which nodes are key players, connectors, or influencers.

Understanding node importance is crucial for many tasks:

• Finding influencers: Who are the most connected people in a social network?
• Identifying critical infrastructure: Which power stations are vital for the grid?
• Stopping disease spread: Which individuals are super-spreaders in an epidemic?

One of the most famous centrality algorithms is PageRank. You might know it from Google!

Developed by Larry Page and Sergey Brin at Stanford University, PageRank was originally used to rank web pages in search results based on their link structure.

Imagine each link from one web page to another is a "vote" of importance. The more links a page receives, the more important it seems.

But not all votes are equal! A vote from an "important" page counts more than a vote from an unimportant one. This creates a recursive process.

PageRank also includes a damping factor. This models a "random surfer" who might get bored and jump to any random page, rather than strictly following links.

Typically set around 0.85, the damping factor ensures that even pages with no incoming links (or "dead ends") can still have some importance, preventing scores from dropping to zero.

Beyond web pages, PageRank is powerful for analyzing social networks.

• Influencer Detection: Find users who are frequently linked to or mentioned by other important users.
• Content Recommendation: Suggest articles or profiles that are highly referenced within a community.

It helps uncover hidden influence patterns.

Consider a tiny network of 3 people: Alice, Bob, and Carol.

• Alice links to Bob and Carol.
• Bob links to Alice.
• Carol links to Alice.

Alice receives votes from Bob and Carol, while also voting for them. After calculation, Alice would likely have a higher PageRank score, indicating her central role in this small communication flow.

Neo4j's Graph Data Science (GDS) Library makes running PageRank easy.

You don't need to write the complex algorithm yourself. GDS provides optimized functions to compute PageRank scores on your graph data directly within Neo4j.

This allows you to quickly identify influential nodes in your datasets.

Great job! You've learned about:

• Centrality algorithms: How they measure node importance in a graph.
• PageRank: Its origin, the "voting" mechanism, and the role of the damping factor.
• Applications: How PageRank identifies influence in various networks.

Next, explore other graph algorithms!