Understand Gu Ge to search engine work principle to do good website to optimize

Core clew: Want to found the website that one is accorded with or gets used to search engine, we need to understand the working principle that searchs engine roughly first. Engine of the following search all is with Google exemple.

Google uses two detector to come the content on capture website: Freshbot and Deepbot. Deepness detector (Deepbot) hits out every months, suffer visit content to be in the main index of Google. Refresh detector (Freshbot) is the content with continual new discovery, for example guest of new website, forum, rich waits. It seems that, google was to discover a new webpage, later often revisit, will look to still have the updating with new what. If have, this new website can be joined have a visit in the list of refresh detector.

The result that refresh detector obtains is collect is in an alone database. When detector of every time refresh has round of new circulation, be rescripted. The main index of refresh detector and Google closes together to provide search result.

The operation mode of Google

Collect—-> collect make up / index—-> the working program of feedback. In fact, search engine includes the following elements.

Capture condition: Search engine expedites detector to arrive on Internet indefatigable collect webpage.

Webpage storehouse: The webpage that the search comes to should be centered in memory of a place, await index processing.

Index arranges: webpage classify, undertake compression, wait undertake indexing making up kind, and the primitive webpage data that did not compress is deleted.

Index condition: the webpage after compressing make a catalogue is under different index.

Ask ask state: The computer language that changes the colloquialism that the user asks ask to place is used into the search what engine reads to understand, will seek advice from each index to get relevant result.

Rank condition: Search index props up will relevant answer to arrange an user with the form of list according to certain level. Search engine thinks best answer is recommended to be in the first place, relatively second arrange subsequently, with this analogize.

SEO is converse search inference

The website should know to search engine above all is how to decide of website quality, this estimate search indexes the process that prop up is converse inference. This converse inference is from the search that searchs engine the rank begins, go exploring search engine to be able to arrange a few websites the front row that searching a result after all, why a few websites do not have even indication chance. The verdict that ultimate deduction gives is what to know to search engine decides and rank a website to existing common law, what knowing the design of the website and promotion are medium then should be done, what should be done.

The PageRank(PR of Google)

PageRank is not is page grade, however admire strange grade. Because the Page here shows is one of author of Google,pull Lipeiji (Larry Page) , can seeing from the name some people are natural is for Internet existence.


PR(A) = (1-d) + D(PR(Ti)/C(Ti)+ . . . + PR(Tn)/C(Tn) )

PR(A) : The admire strange grade of webpage A (PR)

PR(Ti) : The admire strange grade of the webpage Ti that links webpage A (PR) . I can be 0-n, n is the gross of the link. This link can be come from any websites guide link

C(Ti) : The amount that webpage Ti links toward other website (derive link)

D: Damp coefficient, interpose between 0-1, google is set for 0.85.

Can see, the PR of webpage A is decided by the place of other webpage Ti that links it. If there is a link to point to webpage A in webpage Ti, so A got one polls. But if webpage Ti derives mediumly if the link is more, click the possibility that points to A smaller. So the PR of webpage Ti needs to increase advantageous position with C(Ti) only average.

Of course, deciding what the website ranks is PR value not just, still have other factors. But PR is having very main effect in quality of Google judgement website.