The revolutionary ranking of search engines ranks user engagement results

Re-indexing the Internet Facebook hired Google to print public relations process has been completed. The question is: Google so many products, Facebook Why Social Circle so sensitive?

Google: Index Internet

Google claims its mission is to "index the Internet."

Instead of sending as many reptiles, the difficulty of doing this is sorting out the bulk of what's collected: how do you get really important pages ahead of Google's search results?

Google's ranking of search results is aided and manually tuned by other algorithms, but his basic core rule is Pagerank, whose idea is not complicated: "Web pages cited by many premium pages are still premium pages."

field

Pagerank is revolutionary. Another revolution seems to have been overlooked by us:.

Semantic network flirting long after no fly, a simple tag can finally give us something, let us distinguish this page is about which aspects. Since then "area" has also been introduced into the network.

Specific examples of "fields" are Quora or Stack Overflow Tags. Is the problem itself a "technology" or "design" area, a "J2EE" or an "Objcet-C" area?

There is no clear answer to some areas that are not tagged. But there are already people trying to solve these problems. For example, Douban nine, he is to rely on algorithms to automatically blog classification.

expert

The disruptive meaning of sites such as Quora and Stack Overflow is that he has both the concept of "person" and "domain."

The combination of "domain" and "person" produces an "expert": a "person" who specializes in one "domain."

Unlike bloggers who talk nonsense at the time of the blogosphere, real-name webs with Q & A sites like Quora allow us to test a person's profession in one area without the need of a title, certificate, or education Level: The number of questions answered in a particular field is very high, and the rating is particularly good. It is likely to be an expert.

The "experts" on the web are more than just concepts. Some websites have reached the threshold of distinguishing experts. Although known as a copy of Quora, he has a cooler feature than Quora: When inviting people to answer questions, the system recommends possible users.

The underlying algorithm is not hard to guess: Use the question's "fields" to match "people" who answered similar questions. Answering similar questions and gaining a majority of them may well be suitable for answering this question.

Stack Overflow in fact there is a similar evaluation mechanism, called Reputation (reputation). Unfortunately, there is no detail in the field (who has the highest reputation in the question mark Tag Object-C)? However, Stack Overflow want to get this data, is also a matter of minutes.

Vote by experts

Some students may return to ask: This and reindex the Internet have anything to do? We also need to pull down another Facebook features: "Like."

From the above point of view, this magical button is actually a "human" voter on the "web page." If this person also logs on to Quora with a Facebook ID, then this button is a "voter in a field" (relying on Quora's data resolution) for web pages. ("Like" ancestors: Twitter RT too.)

Facebook's new feature "Send" for the non-public sharing? But more importantly: "Like" plus "Send" will allow Facebook to more fully gather a "vote" of people for a "web page" that was previously privately shared via Email, which Facebook could not get.

This shows that the lack of real-name Google +1 is simply an application.

User aggregation and identification

The problem here is that Quora, Stack Overflow, and Twitter are not part of Facebook.

However, do not forget Facebook Connect. If a user logs in to Quora via Facebook ID, it is much easier to determine who is the field expert. And if Facebook ID and Twitter ID can be aggregated, then he likes what, RT and what can be a simple sum.

You should think of FriendFeed here. It seems that all people say: Facebook is FriendFeed acquired for the team. But I have been unable to believe, so valuable team, will find something less important to do?

The above should explain how important it is to aggregate and identify users across social networks.

Re-index the Internet

Words of this, have figure poor dagger see it?

Pagerank is to let the web page to vote, and the data that Facebook holds can let "person", even "expert" come to vote.

Real-name network, in some areas, the votes of experts can give the web a lot of weight, so that he is even higher in the rankings. (The latest example is "the former chief of defense of the United States, chief of staff," announced Osama bin Laden's dead Twitter page.)

Surely a friend will ask: Most pages of "experts" voting, probably also Pagerank very high, right? Probably so. However, long-time observation of the spread of the blog let me note: Facebook Like and Twitter RT accumulation, far more than the reference to the rapid.

In other words, in today's stressed real-time network, social search results may not only be better than Pagerank, but also may be faster.

Foundation of the Internet

Today, as long as a web application wants to do part of the map, he first comes up with the Google Maps API. When most web services use the Google Maps API, Google Mpas is no longer a web-only user-facing application. He precipitated as one of the network's infrastructure.

The importance of Maps, from the comment circle to Google itself, has a clear understanding. So, Google spent a lot of energy scanning the street scene, trying to shoot indoor, or even launch their own satellite. (Google IO two days ago confirmed the following idea: After the successful "indexing the Internet", what Google has to do is "index the real world.")

And Facebook also has the opportunity to become another part of the Internet: By opening up APIs to other applications, the real relationships that he holds can settle down and become another foundation of the web. (Social search is only an application based on this.)

Last words

It seems that Mark Zuckerberg has always assumed the mission of new search and has been pushing Facebook to become a foundation.

Buy FriendFeed; face to face with "Like" to copy Twitter's "RT"; with Facebook Connect to their own platform and so on, all in a block of miniature puzzles.

And Google's Social Circle is the same idea: aggregating and identifying one's identity across different social networks. Later, these data are likely to be used to optimize Google search, or for other web services. Google +1 is an unintelligible product, but what if he can tie in with Social Circle?

Perhaps, these are the real reasons for Facebook to chill.

Finally, the friendship reminds the reader that the above is not a fact, but an individual's guess based on the existing report. Perhaps, I put too much expectations on Facebook; perhaps, so still underestimated their ambitions.

After all, Facebook executives mentioned a slogan within them:

"We just completed a 1% mission."

by Li Nan