How did Larry Page and Sergey Brin's original PageRank algorithm work, and why was it revolutionary?

Okay, let me tell you about this.

Imagine the entire internet as a vast network of academic paper citations.

How does PageRank work?

1. Core Idea: Voting If webpage A links to webpage B, it's like webpage A casting a vote for webpage B. The idea is simple: the more a webpage is linked to, the more important it likely is, just as a paper cited more often is considered more significant.

2. Key Innovation: The "Value" of a Vote Differs This is where Page and Brin's genius truly shone. They realized that not all votes are equal. A vote cast by an "important" webpage holds more weight than a vote from an "unimportant" one.

   For example:

   • A link from the People's Daily website to your personal blog is like a top authority endorsing you.
   • A link from an unknown, newly registered website to your personal blog carries much less weight.

   So, a webpage's "importance" (its PageRank value) depends not only on how many votes it receives but, more crucially, on the "quality" of those votes—that is, how important the webpages casting those votes are themselves. This creates a positive feedback loop: authoritative websites are authoritative because many other websites (especially other authoritative ones) link to them.

Why was it revolutionary?

Before PageRank, early search engines (like AltaVista, Yahoo!) primarily determined a webpage's relevance to a search query by looking at the content of the page itself. They would analyze how many times a word appeared on the page, whether it was in the title, and so on.

The drawback of this method was obvious: it was too easy to manipulate.

Back then, webmasters would aggressively stuff keywords onto their pages to improve rankings. For instance, if you searched for "travel," some pages might be filled with "travel, travel, travel," but the content quality would be terrible, perhaps even a spam page. This led to a very poor user experience, as the search results often weren't what users were looking for.

PageRank's revolution lay in its first-time introduction of "relationships between webpages" to assess a page's quality and authority. It no longer just listened to what a webpage "said about itself" (keyword frequency) but instead listened to "what others said about you" (link votes).

This mechanism simulated how "prestige" and "trust" are transmitted in human society; it trusted the collective intelligence of the entire internet. It was very difficult for a webpage to manipulate its way into getting numerous links from "high-quality websites." Therefore, the results filtered by PageRank offered significantly higher quality and relevance than its competitors at the time.

To summarize simply:

• Before: Search engines were like naive librarians, only checking if the book title and table of contents contained your desired words.
• After (Google): Search engines became intelligent scholars, not only reading the book's content but also seeing how many other "eminent" scholars recommended and cited it.

It was this seemingly simple yet profoundly impactful change that allowed Google to sift gold from a heap of junk information, providing an unprecedented high-quality search experience and ultimately establishing its dominant position in the search industry.